Infectious Disease Early Warning Method and Device, Storage Medium and Electronic Device

Abstract

The present disclosure relates to the field of data mining technologies, and provides an infectious disease early warning method and apparatus, a computer-readable storage medium, and an electronic device. Among them, the above method includes: obtaining a new case number sequence corresponding to a first infectious disease determined based on a preset sampling frequency within the historical transmission duration of the first infectious disease; obtaining a plurality of feature sequences corresponding to a candidate feature in a plurality of candidate time periods, the duration of the candidate time period being the same as the duration of the historical transmission duration, and the time difference between the start date of the candidate time period and the start date of the historical transmission duration being within a first preset value; respectively calculating the correlation degree between the feature sequence corresponding to the candidate feature in each candidate time period and the new case number sequence; and selecting a target feature for early warning of a second infectious disease according to the plurality of correlation degrees corresponding to the plurality of candidate time periods, so as to perform early warning of infectious diseases according to the target feature. The present disclosure can improve the accuracy and timeliness of infectious disease early warning.

Description

Background Art

[0002] Infectious diseases are characterized by fast transmission speed and wide spread. The spread of infectious diseases has a great impact on people's livelihood and the economy. Therefore, early detection, early warning, and early protection of infectious diseases can, to a certain extent, weaken the spread of infectious diseases.

[0003] In the related art, based on the change in the number of hospital admissions in a certain area during a certain period, combined with various warning algorithms, it can be determined whether there is an outbreak of infectious diseases, so as to achieve the purpose of early warning. Taking influenza as an example, the moving percentile method can be used to model the number of patients admitted at different times, and different percentile P values are used as candidate warning critical values to construct an influenza model to achieve the purpose of early warning.

[0004] However, using the number of hospital admissions as a warning feature also means that the number of hospital admissions has reached a certain scale. In this case, the infectious disease may have spread for some time, so there is a certain lag in the early warning of infectious diseases.

[0005] It should be noted that the information disclosed in the above background art section is only used to enhance the understanding of the background of the present disclosure, and thus may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] The purpose of the present disclosure is to provide a method and device for warning of infectious diseases, a computer-readable storage medium, and an electronic device, so as to at least improve the problem that the warning of infectious diseases is not timely to a certain extent.

[0007] Other characteristics and advantages of the present disclosure will become apparent through the following detailed description, or be learned in part through the practice of the present disclosure.

[0008] According to a first aspect of the present disclosure, there is provided a method for warning of infectious diseases, including: obtaining a sequence of new cases corresponding to a first infectious disease determined based on a preset sampling frequency during a historical transmission duration of the first infectious disease; obtaining a plurality of feature sequences corresponding to a candidate feature in a plurality of candidate periods, the duration of each candidate period being the same as the duration of the historical transmission duration, and the time difference between the start date of each candidate period and the start date of the historical transmission duration being within a first preset value, and each feature sequence corresponding to a candidate period being determined according to feature values collected based on the preset sampling frequency during the candidate period; respectively calculating a first degree of correlation between the feature sequences corresponding to the candidate feature in each candidate period and the sequence of new cases to obtain a plurality of first degrees of correlation corresponding to the plurality of feature sequences; and selecting a target feature for warning a second infectious disease from the candidate features according to the plurality of first degrees of correlation, so as to warn the second infectious disease according to the target feature.

[0009] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the selecting, from the candidate features, target features for warning of the second infectious disease according to the multiple first correlation degrees includes: determining the maximum first correlation degree corresponding to the candidate features from the multiple first correlation degrees corresponding to the multiple feature sequences of the candidate features; determining a first virtual score of the candidate features according to the maximum first correlation degree corresponding to the candidate features; and selecting, from the candidate features, target features for warning of the second infectious disease according to the first virtual score.

[0010] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the selecting, from the candidate features, target features for warning of the second infectious disease according to the multiple first correlation degrees includes: obtaining a first cumulative value of the feature values of the candidate features collected based on the preset sampling frequency within a first target time period, where the end date of the first target time period is the start date of the historical propagation duration or a date before the start date of the historical propagation duration, and the duration of the first target time period is a second preset value; determining a first preset number of historical same - period time periods corresponding to the first target time period, and collecting the feature values of the candidate features based on the preset sampling frequency within each historical same - period time period; respectively determining second cumulative values of the feature values of the candidate features collected within each historical same - period time period, to obtain the first preset number of second cumulative values; determining a second virtual score of the candidate features according to the first cumulative value and the first preset number of second cumulative values; and selecting, from the candidate features, target features for warning of the second infectious disease based on the multiple first correlation degrees and the second virtual score.

[0011] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the determining a second virtual score of the candidate features according to the first cumulative value and the first preset number of second cumulative values includes: determining an average value of the first preset number of second cumulative values; determining an absolute value of the difference between the first cumulative value and the average value; and calculating a ratio between the absolute value and the average value to determine the second virtual score of the candidate features.

[0012] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the selecting, from the candidate features, target features for warning of the second infectious disease based on the multiple first correlation degrees and the second virtual score includes: determining the maximum first correlation degree corresponding to the candidate features from the multiple first correlation degrees corresponding to the multiple feature sequences of the candidate features; and determining a first virtual score of the candidate features according to the maximum first correlation degree corresponding to the candidate features.

[0013] Determine the first product of the first virtual score and the first weight corresponding to the first virtual score, and the second product of the second virtual score and the second weight corresponding to the second virtual score respectively; determine the first target virtual score of the candidate feature according to the sum of the first product and the second product; based on the first target virtual score, select the target feature for warning of the second infectious disease from the candidate features.

[0014] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the selecting the target feature for warning of the second infectious disease from the candidate features based on the plurality of first correlation degrees and the second virtual score includes: taking a preset time interval as the target period, and determining a second preset number of target periods before the historical propagation duration period; sampling the candidate features in each of the target periods according to the preset sampling frequency to respectively determine the to-be-compared feature sequences corresponding to the candidate features in each of the target periods; sampling the candidate features in any one of the target periods within the historical propagation duration period based on the preset sampling frequency to obtain the target feature sequence of the candidate features; determining the third virtual score of the candidate feature according to the difference degree between the to-be-compared feature sequence and the target feature sequence; selecting the target feature for warning of the second infectious disease from the candidate features based on the plurality of first correlation degrees, the second virtual score, and the third virtual score.

[0015] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the selecting the target feature for warning of the second infectious disease from the candidate features according to the plurality of first correlation degrees includes: taking a preset time interval as the target period, and determining a second preset number of target periods before the historical propagation duration period; sampling the candidate features in each of the target periods according to the preset sampling frequency to respectively determine the to-be-compared feature sequences corresponding to the candidate features in each of the target periods; sampling the candidate features in any one of the target periods within the historical propagation duration period based on the preset sampling frequency to obtain the target feature sequence of the candidate features; determining the third virtual score of the candidate feature according to the difference degree between the to-be-compared feature sequence and the target feature sequence; selecting the target feature for warning of the second infectious disease from the candidate features based on the plurality of first correlation degrees and the third virtual score.

[0016] In an exemplary embodiment of the present disclosure, based on the foregoing solution, selecting target features for warning of a second infectious disease from the candidate features based on the multiple first correlation degrees and the third virtual score includes: determining the maximum first correlation degree corresponding to the candidate feature from the multiple first correlation degrees corresponding to the multiple feature sequences of the candidate feature; determining the first virtual score of the candidate feature according to the maximum first correlation degree corresponding to the candidate feature; respectively determining the first product of the first virtual score and the first weight corresponding to the first virtual score, and the third product of the third virtual score and the third weight corresponding to the second virtual score; determining the second target virtual score of the candidate feature according to the sum of the first product and the third product; and selecting target features for warning of a second infectious disease from the candidate features based on the second target virtual score.

[0017] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the difference degree is determined by the following method: calculating the second correlation degree between each of the to-be-compared feature sequences and the target feature sequence respectively to obtain the second preset number of second correlation degrees; and determining the difference degree between the to-be-compared feature sequence and the target feature sequence according to the mean value of the second preset number of second correlation degrees based on the negative correlation relationship.

[0018] In an exemplary embodiment of the present disclosure, based on the foregoing solution, selecting target features for warning of a second infectious disease from the candidate features based on the multiple first correlation degrees, the second virtual score, and the third virtual score includes: determining the maximum first correlation degree corresponding to the candidate feature from the multiple first correlation degrees corresponding to the multiple feature sequences of the candidate feature; determining the first virtual score of the candidate feature according to the maximum first correlation degree corresponding to the candidate feature; respectively determining the first product of the first virtual score and the first weight corresponding to the first virtual score, the second product of the second virtual score and the second weight corresponding to the second virtual score, and the third product of the third virtual score and the third weight corresponding to the third virtual score; calculating the sum of the first product, the second product, and the third product to determine the third target virtual score of the candidate feature in the first infectious disease; and selecting target features for warning of a second infectious disease from the candidate features based on the third target virtual score.

[0019] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the second weight and the third weight are determined in the following manner: The candidate features are sorted according to the descending order of the maximum first correlation degree corresponding to the candidate features, and the first N reference candidate features are determined, where N is a positive integer; for each reference candidate feature, the first error between the maximum first correlation degree of the reference candidate feature and the second virtual score of the reference candidate feature, and the second error between the maximum first correlation degree of the reference candidate feature and the third virtual score of the reference candidate feature are calculated respectively; according to the first error and the second error corresponding to each reference candidate feature, the second weight and the third weight are determined.

[0020] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the determining the second weight and the third weight according to the first error and the second error corresponding to each reference candidate feature includes: calculating the first error mean of the first errors corresponding to the N reference candidate features and the second error mean of the second errors corresponding to the N reference candidate features respectively; generating an input vector according to the opposite number of the first error mean and the opposite number of the second error mean, and inputting the input vector into a probability mapping function to obtain a first probability corresponding to the second virtual score and a second probability corresponding to the third virtual score; determining the second weight corresponding to the second virtual score according to the first probability, and determining the third weight corresponding to the third virtual score according to the second probability.

[0021] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the first infectious disease and the second infectious disease are the same; the selecting the target features for warning the second infectious disease from the candidate features based on the third target virtual score includes: sorting the candidate features in descending order based on the third target virtual score; determining the target features for warning the first infectious disease according to the first M candidate features in the sorting result, where M is a positive integer.

[0022] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the first infectious disease and the second infectious disease are different; the selecting the target features for warning the second infectious disease from the candidate features based on the third target virtual score includes: obtaining the mean value of the third target virtual scores corresponding to the candidate features in multiple different first infectious diseases; sorting the candidate features in descending order according to the mean value of the third target virtual scores; determining the target features for warning the second infectious disease according to the first M candidate features in the sorting result, where M is a positive integer.

[0023] In an exemplary embodiment of the present disclosure, based on the foregoing solution, the first correlation degree is determined based on any one of the Pearson correlation coefficient, the T-test, and the rank sum test.

[0024] According to a second aspect of the present disclosure, there is provided an infectious disease early warning device, including: a newly added number sequence determination module configured to obtain, during a historical transmission duration of a first infectious disease, a newly added number sequence corresponding to the first infectious disease determined based on a preset sampling frequency; a plurality of feature sequence acquisition modules configured to obtain a plurality of feature sequences corresponding to a candidate feature in a plurality of candidate time periods, the duration of each candidate time period being the same as the duration of the historical transmission duration, and the time difference between the start date of each candidate time period and the start date of the historical transmission duration being within a first preset value, and the feature sequence corresponding to each candidate time period being determined according to feature values collected based on the preset sampling frequency during the candidate time period; a first correlation degree calculation module configured to calculate respectively the first correlation degree between the feature sequence corresponding to the candidate feature in each candidate time period and the newly added number sequence, so as to obtain a plurality of first correlation degrees corresponding to the plurality of feature sequences; and an infectious disease early warning module configured to select, according to the plurality of first correlation degrees, a target feature for early warning a second infectious disease from the candidate features, so as to early warn the second infectious disease according to the target feature.

[0025] According to a third aspect of the present disclosure, there is provided a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the infectious disease early warning method as described in the first aspect in the above embodiment.

[0026] According to a fourth aspect of the embodiments of the present disclosure, there is provided an electronic device, including: a processor; and a storage device for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the infectious disease early warning method as described in the first aspect in the above embodiment.

[0027] As can be seen from the above technical solutions, the infectious disease early warning method, the infectious disease early warning device, the computer-readable storage medium and the electronic device for implementing the infectious disease early warning method in the exemplary embodiments of the present disclosure at least have the following advantages and positive effects:

[0028] In the technical solutions provided by some embodiments of the present disclosure, it is possible to obtain a sequence of the number of new cases corresponding to the first infectious disease determined based on a preset sampling frequency during the historical transmission duration of the first infectious disease. It is also possible to generate a plurality of candidate time periods according to the scanning phase difference. Specifically, the duration of each candidate time period is the same as the duration corresponding to the historical transmission duration, and the time difference between the start date of each candidate time period and the start date of the historical transmission duration is within a first preset value. Then, a plurality of feature sequences corresponding to the candidate features in the plurality of candidate time periods are obtained, where the feature sequence corresponding to each candidate time period is determined according to the feature values collected based on the preset sampling frequency during the candidate time period. Next, the first correlation degree between the feature sequence corresponding to the candidate feature in each candidate time period and the sequence of the number of new cases is calculated respectively to obtain a plurality of first correlation degrees corresponding to the plurality of feature sequences. Thus, according to the plurality of first correlation degrees, a target feature for warning the second infectious disease is selected from the candidate features, and an early warning of the infectious disease is made based on the target feature. Compared with the related art, on the one hand, the present disclosure determines a plurality of candidate feature sequences based on the generated plurality of candidate time periods, and there is a certain phase difference between the plurality of candidate time periods and the historical transmission duration. Therefore, warning features related to the lag or lead of the transmission of the infectious disease can be accurately determined, thereby improving the accuracy of the infectious disease early warning. On the other hand, the present disclosure can avoid the problem of warning lag caused by directly using only the number of new cases of the infectious disease as the warning index by analyzing the correlation between other candidate features and the sequence of the number of new cases of the infectious disease, and can improve the timeliness of the infectious disease early warning.

[0029] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure. Obviously, the accompanying drawings in the following description are only some embodiments of the present disclosure. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0031] Figure 1 A flowchart showing the process of the infectious disease early warning method in an exemplary embodiment of the present disclosure;

[0032] Figure 2 A flowchart showing the process of the method for determining the first virtual score in an exemplary embodiment of the present disclosure;

[0033] Figure 3 A flowchart showing the process of the method for determining the second virtual score in an exemplary embodiment of the present disclosure;

[0034] Figure 4 A flowchart showing a method for determining a third virtual score in an exemplary embodiment of the present disclosure;

[0035] Figure 5 A flowchart showing a method for warning of emerging infectious diseases in an exemplary embodiment of the present disclosure;

[0036] Figure 6 A flowchart showing a method for determining a second weight and a third weight in an exemplary embodiment of the present disclosure;

[0037] Figure 7 A schematic structural diagram of an infectious disease warning device in an exemplary embodiment of the present disclosure;

[0038] Figure 8 A schematic structural diagram of an electronic device in an exemplary embodiment of the present disclosure. Detailed implementation manners

[0039] Example embodiments will now be described more fully with reference to the accompanying drawings. However, the example embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be more thorough and complete, and will fully convey the concept of the example embodiments to those skilled in the art. The features, structures, or characteristics described may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of the embodiments of the present disclosure. However, those skilled in the art will recognize that the technical solutions of the present disclosure can be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. may be employed. In other instances, well-known technical solutions are not shown or described in detail to avoid obscuring the various aspects of the present disclosure.

[0040] As used in this specification, the terms "a", "an", "the", and "said" are used to indicate the presence of one or more elements / components / etc.; the terms "comprising" and "having" are used to mean an open inclusion and mean that in addition to the listed elements / components / etc., there may be additional elements / components / etc.; the terms "first", "second", "third", etc. are used only as labels and are not a limitation on the quantity of their objects.

[0041] In addition, the drawings are only schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus their repeated description will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities.

[0042] Infectious diseases refer to infectious diseases caused by pathogenic microorganisms, which are usually characterized by rapid transmission and wide prevalence. The spread of infectious diseases has a great impact on people's livelihood and the economy. Early detection, early warning and early prevention of infectious diseases will greatly reduce the spread of infectious diseases. Therefore, early warning of infectious diseases is crucial.

[0043] In related technologies, it is possible to directly determine whether there is an outbreak of infectious diseases based on the changes in the number of hospitalizations in a certain area during a certain period of time, combined with various early warning algorithms, thereby achieving the purpose of early warning.

[0044] Taking influenza as an example, the moving percentile method can be used to model the number of patients admitted to the hospital at different times, and different percentiles P can be used as candidate warning critical values ​​to construct an influenza model to achieve the purpose of warning.

[0045] However, different people may choose different measures after being infected with an infectious disease, such as staying at home, taking medicine, or seeking medical treatment. In order to avoid false alarms, different early warning models will often issue early warnings only after the number of hospitalizations reaches a certain trend, large or small. In this case, the infectious disease may have spread for a certain period of time and reached a certain scale of infection, so there is a certain lag in the early warning of the infectious disease, and the control of the infectious disease cannot achieve the purpose of timely and effective control.

[0046] In view of one or more of the above problems, in an embodiment of the present disclosure, a method for determining infectious disease warning characteristics is first provided, which can mine more indicators or characteristics related to infectious diseases as a reference for infectious disease warning, thereby overcoming the defects existing in the above-mentioned related technologies to a certain extent and achieving the purpose of more accurate and timely early warning of infectious diseases.

[0047] Figure 1 A flow chart showing a method for determining early warning characteristics of an infectious disease in an exemplary embodiment of the present disclosure is shown. Figure 1 , the method may include:

[0048] Step S110, obtaining a sequence of newly added people corresponding to the first infectious disease determined based on a preset sampling frequency during a historical transmission period of the first infectious disease;

[0049] Step S120, obtaining multiple feature sequences corresponding to the candidate features in multiple candidate time periods, the duration of each candidate time period is the same as the duration corresponding to the historical propagation duration period, and the time difference between the start date of each candidate time period and the start date of the historical propagation duration period is within a first preset value, and the feature sequence corresponding to each candidate time period is determined according to the feature value collected in the candidate time period based on the preset sampling frequency;

[0050] Step S130: Calculate the first correlation degree between the feature sequence corresponding to each candidate time period of the candidate feature and the new - added population sequence respectively, so as to obtain multiple first correlation degrees corresponding to multiple feature sequences.

[0051] Step S140: Select a target feature for warning against the second infectious disease from the candidate features according to the multiple first correlation degrees, so as to warn against the second infectious disease according to the target feature.

[0052] In Figure 1 In the technical solution provided by the embodiment shown, it is possible to obtain the new - added population sequence corresponding to the first infectious disease determined based on a preset sampling frequency during the historical transmission duration of the first infectious disease. It is also possible to generate multiple candidate time periods according to the scanning phase difference. Specifically, the duration of each candidate time period in the multiple candidate time periods is the same as the duration corresponding to the historical transmission duration, and the time difference between the start date of each candidate time period and the start date of the historical transmission duration is within a first preset value. Then, obtain multiple feature sequences corresponding to the candidate feature in multiple candidate time periods, where the feature sequence corresponding to each candidate time period is determined according to the feature values collected based on the preset sampling frequency during the candidate time period. Then, calculate the first correlation degree between the feature sequence corresponding to each candidate time period of the candidate feature and the new - added population sequence respectively, so as to obtain multiple first correlation degrees corresponding to multiple feature sequences. Thus, select a target feature for warning against the second infectious disease from the candidate features according to the multiple first correlation degrees, so as to warn against the infectious disease according to the target feature. Compared with the related technology, on the one hand, based on the generated multiple candidate time periods, this disclosure determines multiple candidate feature sequences. Since there is a certain phase difference between the multiple candidate time periods and the historical transmission duration, it is possible to accurately determine the warning features related to the lag or lead in the transmission of the infectious disease, improving the accuracy of infectious disease warning. On the other hand, by analyzing the correlation between other candidate features and the new - added population sequence of the infectious disease, this disclosure can select other warning features, avoiding the problem of warning lag caused by directly using only the new - added population of the infectious disease as the warning feature, and improving the timeliness of infectious disease warning.

[0053] The following Figure 1 elaborates on the specific implementation manners of each step in the embodiment shown in detail:

[0054] In step S110, obtain the new - added population sequence corresponding to the first infectious disease determined based on a preset sampling frequency during the historical transmission duration of the first infectious disease.

[0055] In an alternative embodiment, the historical transmission duration period may include one or more, for example, it may include the transmission duration period in the target area when the first infectious disease occurred most recently, or occurred at any time in the past, or occurred every time or several times recently. Among them, the start date of the historical transmission duration period may include the date when the first infectious disease first appeared, and the end date of the transmission duration period may include the date when the infectious disease was effectively controlled and no longer continued to spread.

[0056] The scope of the target area can be customarily determined according to one or more administrative regions or geographical regions. For example, the target area can be the world, Asia, China, the central region of China (including the three provinces of Henan, Hubei, and Hunan), or a certain city in China such as Beijing, or a certain urban area of a certain city, such as a certain district in Beijing, etc. This exemplary embodiment does not make special limitations on this.

[0057] After determining the historical transmission duration period of the first infectious disease in the target area, the sequence of the number of newly added cases of the first infectious disease within the historical transmission duration period in the target area can be determined based on a preset sampling frequency. Exemplarily, according to the preset sampling frequency, sampling can be performed on the records of the number of newly added cases of the first infectious disease in the statistically historical transmission duration period of the target area to obtain the sequence of the number of newly added cases corresponding to the first infectious disease.

[0058] Among them, the records of the number of newly added cases can be obtained through channels such as hospitals and / or other institutions related to infectious disease prevention, such as the Centers for Disease Control and Prevention. The preset sampling frequency can be customarily determined according to requirements. For example, the preset sampling frequency can be daily. In the records of the number of newly added cases of the first infectious disease, the number of newly added cases of the first infectious disease per day within the historical transmission duration period can be collected, so as to obtain the sequence of the number of newly added cases of the first infectious disease per day within the historical transmission duration period.

[0059] As mentioned above, the historical duration period may include one or more, and the corresponding sequence of the number of newly added cases may also include one or more. Taking the target area as a certain district in Beijing as an example, the historical transmission duration period of a certain infectious disease in a certain district can be shown in Table 1. Among them, each outbreak period in Table 1 can be understood as a historical transmission duration period, that is, the infectious disease has broken out 3 times in a certain district in Beijing. The sequence of the number of newly added cases within each outbreak period can be obtained based on the preset sampling frequency. For example, the sequence of the number of newly added cases per day 1 corresponding to outbreak period 1, the sequence of the number of newly added cases per day 2 corresponding to outbreak period 2, and the sequence of the number of newly added cases per day 3 corresponding to outbreak period 3 can be obtained respectively.

[0060] Table 1 Historical outbreak periods of a certain infectious disease

[0061] Region Outbreak Period 1 Outbreak Period 2 Outbreak Period 3 A certain district in Beijing From July 1, 2013 to September 7, 2013 From January 1, 2012 to March 7, 2012 From February 2, 2015 to April 6, 2015

[0062] Taking the preset sampling frequency as an example of once a day, for each historical transmission duration period, the information on the number of newly added cases per day during the corresponding historical transmission duration period of the infectious disease in the region can be statistically analyzed to obtain the daily newly added case sequence corresponding to each historical transmission duration period respectively. Table 2 takes a certain district in Beijing as an example and shows the information on the number of newly added cases in some dates during a certain historical transmission duration period of a certain infectious disease. Based on the information on the number of newly added cases and the preset sampling frequency, the newly added case sequence in the historical transmission duration period of the target region can be generated.

[0063] Table 2 Number of newly added cases of a certain infectious disease in a certain historical transmission duration period in a certain district of Beijing

[0064] Region 2020-1-1 2020-1-2 2020-1-3 2020-1-4 2020-1-5 A certain district in Beijing 2 3 10 2 9

[0065] Continue to refer to Figure 1 In step S120, multiple feature sequences corresponding to a candidate feature in multiple candidate time periods are obtained.

[0066] For example, in the present disclosure, features strongly associated with the infectious disease can be obtained by performing search and sorting within a limited feature range. Therefore, it is possible to first determine which potential factors can be used as candidate features to select features strongly associated with the infectious disease from all candidate features.

[0067] In an alternative embodiment, the candidate features may include any features that may be related to the transmission of the infectious disease. For example, they may include weather, traffic, sales information of different drugs, sales information of different commodities, medical information in hospitals, etc. Of course, they may also include other features that may have an impact on the spread of the infectious disease. This exemplary embodiment does not make special limitations on this.

[0068] Among them, the sales information of different drugs or commodities may include the sales quantity of different commodities or drugs. The weather may include temperature, humidity, etc. The medical information in hospitals may include the change in the number of people tested for medical indicators, such as the number of people suffering from influenza in the hospital.

[0069] In an alternative embodiment, the candidate features may include time-varying features or non-time-varying features. Among them, time-varying features can be understood as features whose feature values change with time.

[0070] In an alternative embodiment, for any historical transmission duration period of the first infectious disease, multiple candidate time periods corresponding to the historical transmission duration period can be generated according to the historical transmission duration period and a first preset value. Then, for each candidate feature, multiple feature sequences corresponding to multiple candidate time periods can be respectively obtained to generate multiple feature sequences corresponding to each candidate feature in multiple candidate time periods.

[0071] In an alternative embodiment, the duration of each candidate period among the multiple candidate periods is the same as the duration corresponding to the historical propagation duration period, and the time difference between the start date of each candidate period and the start date of the historical propagation duration period is within a first preset value. The feature sequence corresponding to each candidate period is determined based on the feature values collected at the preset sampling frequency within the candidate period.

[0072] Among them, the first preset value can be custom-determined according to requirements, such as 30 days, 20 days, 50 days, etc., and this exemplary embodiment does not make special limitations on this. The start date of the candidate period can be earlier than the start date of the corresponding historical propagation duration period or later than the start date of the corresponding historical propagation duration period.

[0073] For example, based on the idea of scanning phase difference, the time window corresponding to the historical propagation duration period can be moved forward and backward by the first preset value to generate multiple candidate periods. Taking moving forward and backward by 30 days as an example, if the time window corresponding to the historical propagation duration period is [a, b], where a represents the start date and b represents the end date, then the multiple candidate periods can correspond to the time windows from [a - 30 days, b - 30 days] to [a + 30 days, b + 30 days].

[0074] For each candidate feature, the feature sequence corresponding to it in each candidate period can be obtained respectively. Taking the first preset value as 30 days as an example, 61 candidate periods can be obtained. Among these 61 candidate periods, the historical propagation duration period itself is included. Then, for each candidate feature, 61 feature sequences corresponding to the 61 candidate periods can be obtained.

[0075] Taking the historical propagation duration period itself in the candidate period as an example, the feature values of each candidate feature within the statistically historical propagation duration period can be sampled according to the preset sampling frequency to obtain the feature sequences corresponding to each candidate feature within the historical propagation duration period.

[0076] Taking the preset sampling frequency as daily as an example, for each candidate feature, the feature value of each day within the historical propagation duration period in the target area can be collected in units of days, so as to generate the feature sequence corresponding to each candidate feature within the historical propagation duration period.

[0077] Taking the target area as a certain district in Beijing, the candidate feature as temperature, and the preset sampling frequency as daily as an example, the temperature values of each day in a certain district in Beijing within the historical propagation duration period can be obtained, as shown in Table 3. Based on the obtained temperature values of each day, the feature sequence of the temperature feature within the historical propagation duration period can be generated. Only the temperature values of some dates are shown in Table 3.

[0078] Table 3 Partial eigenvalue of temperature characteristics in a certain district of Beijing during a certain historical propagation duration

[0079] Feature Region 2020-1-1 2020-1-2 2020-1-3 2020-1-4 2020-1-5 Temperature A certain district in Beijing 6.8℃ 6.2℃ 4℃ 2.1℃ 3.2℃

[0080] The acquisition method of the feature sequences corresponding to each candidate feature in other candidate time periods is the same as the specific implementation method of obtaining the feature sequences corresponding to each candidate feature in the historical propagation duration described above, and will not be elaborated here.

[0081] Next, continue to refer to Figure 1 , in step S130, calculate the first correlation degree between the feature sequences corresponding to the candidate feature in each candidate time period and the new population number sequence respectively, so as to obtain multiple first correlation degrees corresponding to multiple feature sequences.

[0082] Continuing with the candidate time period being the historical propagation duration shifted forward and backward by 30 days as described above, resulting in 61 candidate time periods as an example, for each candidate feature, the first correlation degree between the 61 feature sequences corresponding to the above 61 candidate time periods and the new population number sequence obtained in step S110 can be calculated respectively. In this way, for each candidate feature, 61 first correlation degrees can be obtained.

[0083] In an alternative embodiment, the first correlation degree can be determined based on any one of the Pearson correlation coefficient, T-test, and rank sum test. Of course, the first correlation degree can be determined by other methods of calculating the correlation degree between two variables, and this exemplary embodiment does not make special limitations on this.

[0084] Taking the Pearson correlation coefficient as an example, for each candidate feature, the first correlation degree between the feature sequence corresponding to it in each candidate time period and the new population number sequence obtained in step S110 can be calculated respectively based on the Pearson correlation coefficient.

[0085] Among them, the Pearson correlation coefficient, also known as the Pearson product-moment correlation coefficient, is a linear correlation coefficient and is one of the most commonly used correlation coefficients. Denoted as r, it is used to reflect the linear correlation degree between two variables X and Y. The value of r ranges from -1 to 1, and the larger the absolute value, the stronger the correlation. Using the Pearson correlation coefficient, the positive correlation degree and the negative correlation degree can be obtained. For candidate features with strong negative correlation, they can also be considered as features strongly related to infectious diseases.

[0086] After obtaining multiple first correlation degrees corresponding to each candidate feature, in step S140, according to the multiple first correlation degrees, target features for warning against the second infectious disease are selected from the candidate features, so as to warn against the second infectious disease according to the target features.

[0087] In an alternative embodiment, multiple first virtual scores of each candidate feature can be determined according to the multiple first correlation degrees, and then based on the first virtual scores, target features for warning against the second infectious disease are selected from the candidate features.

[0088] Exemplarily, Figure 2 A flowchart showing a method for determining the first virtual score in an exemplary embodiment of the present disclosure is shown. Refer to Figure 2 This method may include steps S210 to S230. Wherein:

[0089] In step S210, the maximum first correlation degree corresponding to the candidate feature is determined from the multiple first correlation degrees corresponding to the multiple feature sequences of the candidate feature.

[0090] In an alternative embodiment, when the historical propagation duration in step S110 is one, for each candidate feature, a maximum first correlation degree is directly determined from the multiple first correlation degrees.

[0091] In an alternative embodiment, when the historical propagation duration in step S110 is multiple, for each candidate feature, it can respectively determine the maximum first correlation degree from the multiple first correlation degrees of the multiple candidate time periods corresponding to each historical propagation duration, that is, for each candidate feature, multiple maximum first correlation degrees can be determined according to the number of historical propagation durations.

[0092] Next, in step S220, the first virtual score of the candidate feature is determined according to the maximum first correlation degree corresponding to the candidate feature.

[0093] In an alternative embodiment, for each candidate feature, when the number of maximum first correlation degrees determined in step S210 is 1, the maximum first correlation degree can be directly determined as the first virtual score of the candidate feature, or the first virtual score of the candidate feature can be determined based on the positive correlation relationship between the maximum first correlation degree of the candidate feature and the first virtual score of the candidate feature according to the maximum first correlation degree of the candidate feature.

[0094] In an alternative embodiment, for each candidate feature, when there are multiple maximum first correlation degrees determined in step S210, the average value of the multiple maximum first correlation degrees can be used as the first virtual score of the candidate feature. For example, the average value of the multiple maximum first correlation degrees can be directly determined as the first virtual score of the candidate feature, or according to the positive correlation relationship, based on the average value of the maximum first correlation degree of the candidate feature, the first virtual score of the candidate feature can be determined.

[0095] In the present disclosure, based on the idea of scanning phase difference, multiple candidate time periods corresponding to the historical propagation duration are generated according to a first preset value, and then for each candidate feature, multiple candidate feature sequences are determined according to the multiple candidate time periods. Thus, based on the maximum first correlation degree among the multiple first correlation degrees corresponding to the multiple candidate feature sequences, the first virtual score of the candidate feature is determined, and then based on the first virtual score, the target feature for early warning is determined. In this way, some candidate features that are lag-correlated or lead-correlated with the number of infectious disease patients can be accurately screened out, improving the accuracy of determining infectious disease early warning features.

[0096] Next, in step S230, according to the first virtual score, the target feature for warning against the second infectious disease is selected from the candidate features.

[0097] In an exemplary embodiment, the first infectious disease and the second infectious disease may be the same or different. In an alternative embodiment, the target feature for warning against the second infectious disease can be directly selected from the candidate features according to the first virtual score.

[0098] Specifically, when the first infectious disease and the second infectious disease are the same, the candidate features can be sorted according to the descending order of the absolute value of the first virtual score, and the first M candidate features in the sorting result are determined, and the first M candidate features are determined as the target features for warning against the first infectious disease, where M is a positive integer and can be custom-defined according to requirements.

[0099] When the first infectious disease is different from the second infectious disease, the above steps S110 to S140 can be respectively executed for multiple different first infectious diseases to obtain the first virtual scores of each candidate feature in each first infectious disease. Among them, the first virtual score can be determined according to a historical transmission duration, or can be determined according to the average value of the first virtual scores corresponding to multiple historical transmission durations. Then, for each candidate feature, the first virtual scores corresponding to it in each first infectious disease are weighted and averaged to obtain the average value of the first virtual scores of each candidate feature in multiple different first infectious diseases. Then, the candidate features are sorted according to the descending order of this average value, and the top M candidate features in the sorting result are determined as the target features for warning the second infectious disease.

[0100] When the first infectious disease is different from the second infectious disease, the above steps S110 to S140 can also be respectively executed for multiple different first infectious diseases to determine the target features corresponding to each first infectious disease. Then, take the intersection of the target features corresponding to each first infectious disease, that is, the same target features corresponding to each first infectious disease, as the target features for warning the second infectious disease.

[0101] In the present disclosure, after determining the target features for warning the second infectious disease, the second infectious disease can be warned according to the determined target features. Among them, the warning method can be freely selected according to requirements. For example, using the historical feature values of the target features as the input of the training samples and using whether the second infectious disease breaks out as the label of the training samples to train a deep learning model to obtain a warning model for warning the second infectious disease, and then warning the second infectious disease according to the warning model. Of course, it is also possible to warn the second infectious disease based on the target features according to other infectious disease warning methods, and this exemplary embodiment does not make special limitations on this.

[0102] In the present disclosure, when the first infectious disease is different from the second infectious disease, it is possible to realize the warning of newly emerging unknown infectious diseases based on the determined target features.

[0103] In an alternative embodiment, the second virtual score of the candidate feature can also be determined. Based on this, the specific implementation manner of step S140 can also be to select the target features for warning the second infectious disease from the candidate features based on the multiple first correlation degrees and the second virtual score.

[0104] Exemplarily, Figure 3 The flowchart showing the method for determining the second virtual score in an exemplary embodiment of the present disclosure is shown. Refer to Figure 3 and the method may include steps S310 to S340.

[0105] In step S310, a first cumulative value of the feature values ​​of the candidate features collected based on the preset sampling frequency within a first target time period is obtained.

[0106] In an optional implementation, the end date of the first target time period is the start date of the historical communication duration period or a date before the start date of the historical communication duration period, and the duration of the first target time period is a second preset value.

[0107] When the end date of the first target time period is a date before the start date of the historical communication duration period, the time difference between the end date of the first target time period and the start date of the historical communication duration period is less than a preset value, such as less than 2 days, that is, the end date of the first target time period is 1 day before or 2 days before the start date of the historical communication duration period.

[0108] In other words, the end date of the first target time period is at least the start date of the historical communication duration period and cannot be too far away from the start date of the historical communication duration period.

[0109] The length of the first target time period can be determined according to demand, such as 90 days, 60 days, etc., and this exemplary embodiment does not specifically limit this.

[0110] Taking the first target time period as 90 days, the preset collection frequency as daily, and the historical propagation duration as [a, b] as an example, the first target time period may be [a-90, a]. For each candidate feature, its feature value for each day during the period [a-90, a] may be collected to obtain a feature sequence of each candidate feature in the first target time period, and then all feature values ​​in the feature sequence are summed to obtain a first cumulative value of the feature value of each candidate feature in the first target time period.

[0111] Next, in step S320, a first preset number of historical concurrent time periods corresponding to the first target time period is determined, and feature values ​​of the candidate features are collected in each historical concurrent time period based on the preset sampling frequency.

[0112] In an optional implementation, the historical contemporaneous period corresponding to the first target time period can be understood as a period of time that is the same as the start date and end date of the first target time period in a year before the first target time period and in which the spread of the first infectious disease has not occurred. For example, if the first target time period is from December 25, 2020 to March 1, 2021, and the spread of the first infectious disease has not occurred before 2020, then a corresponding historical contemporaneous period can be from December 25, 2019 to March 1, 2020.

[0113] Among them, the first preset quantity can be customized according to requirements, and the first preset quantity is an integer greater than or equal to 1. This exemplary embodiment does not make special limitations on this.

[0114] After determining the historical same - period time periods corresponding to the first target time period, the feature values of each candidate feature can be collected respectively within each historical same - period time period based on the preset sampling frequency. Continuing with the example where the preset sampling frequency is daily, for each candidate feature, the feature values of each day within each historical same - period time period can be collected respectively, and a feature sequence of the candidate feature within the historical same - period time period is obtained.

[0115] In step S330, the second cumulative values of the feature values of the candidate features collected within each historical same - period time period are determined respectively, and the first preset quantity of second cumulative values is obtained.

[0116] Exemplarily, for each candidate feature, the feature values in the feature sequence within each historical same - period time period are added respectively to obtain the second cumulative value of the feature values within each historical same - period time period. In other words, the number of the obtained second cumulative values is the same as the number of historical same - period time periods.

[0117] After obtaining the first cumulative value and the first preset quantity of second cumulative values, in step S340, according to the first cumulative value and the first preset quantity of second cumulative values, the second virtual score of the candidate feature is determined.

[0118] In an alternative embodiment, the specific implementation of step S340 may include: determining the average value of the first preset quantity of second cumulative values; determining the absolute value of the difference between the first cumulative value and the average value; calculating the ratio between the absolute value and the average value to determine the second virtual score of the candidate feature.

[0119] For example, according to the following formula (1), the second virtual score of candidate feature i can be determined based on the first cumulative value and the first preset quantity of second cumulative values.

[0120]

[0121] In formula (1), C(F) represents the second virtual score, abs represents taking the absolute value, mean represents taking the mean, S(F i ) represents the first cumulative value of candidate feature i within the first target time period, and S in represents the second cumulative value of feature i within the nth historical same - period time period.

[0122] Taking the first target time period as [a - 90, a] mentioned above and the first preset quantity as 3 for example, for each candidate feature i, its feature values within [a - 90, a] can be collected according to the preset sampling frequency, and all the collected feature values are accumulated to obtain the first cumulative value S(F i ) of the feature values of the candidate feature i, and then the second cumulative value S i1 , S i2 , S i3 of the feature values of the candidate feature i within the same time periods of the past 3 years without the spread of the first infectious disease is determined. Then, the second virtual score of the candidate feature i can be determined as:

[0123] In the present disclosure, when the first preset quantity is multiple, that is, an integer greater than 1, the cumulative value of the candidate feature in the historical same - time periods can be determined according to the average value of the second cumulative values of multiple historical same - time periods, which improves the stability and accuracy of the cumulative value of the feature values of the candidate feature in the historical same - time periods, and further improves the accuracy of the determination of the second virtual score.

[0124] After determining the second virtual score, based on the multiple first correlation degrees and the second virtual score corresponding to each candidate feature, the target feature for warning of the second infectious disease can be selected from the candidate features.

[0125] Exemplarily, the specific implementation manner of selecting the target feature for warning of the second infectious disease from the candidate features based on the multiple first correlation degrees and the second virtual score corresponding to each candidate feature may include: determining the maximum first correlation degree corresponding to the candidate feature from the multiple first correlation degrees corresponding to the multiple feature sequences of the candidate feature; determining the first virtual score of the candidate feature according to the maximum first correlation degree corresponding to the candidate feature; respectively determining the first product of the first virtual score and the first weight corresponding to the first virtual score, and the second product of the second virtual score and the second weight corresponding to the second virtual score; determining the first target virtual score of the candidate feature according to the sum of the first product and the second product; and then, based on the first target virtual score, selecting the target feature for warning of the second infectious disease from the candidate features.

[0126] For example, for the multiple first correlation degrees corresponding to each candidate feature, the maximum first correlation degree among the multiple first correlation degrees can be determined. Then, based on the maximum first correlation degree corresponding to each candidate feature, the first virtual score of the candidate feature can be determined. Then, according to the first weight and the second weight, the first virtual score and the second virtual score of each candidate feature can be weighted and summed, and the result of the weighted sum can be used as the first target virtual score of each candidate feature. Then, the target feature is selected according to the first target virtual score of each candidate feature.

[0127] Among them, the specific implementation manner of selecting the target feature according to the first target virtual score is the same as the specific implementation manner of selecting the target feature according to the first virtual score above. Just replace the technical term "first virtual score" with "first target virtual score", and details will not be elaborated here.

[0128] Similarly, when the historical transmission duration of a certain first infectious disease includes multiple durations, the above steps can be respectively executed according to each historical transmission duration to respectively obtain the first target virtual scores of each feature in each historical transmission duration. Then, the first target virtual scores corresponding to each historical transmission duration are weighted and averaged, and the final first target virtual scores of each candidate feature in the first infectious disease are determined according to the weighted average first target virtual scores.

[0129] In practice, the feature sequences of some candidate features may not be highly correlated with the transmission of infectious diseases, but the cumulative effects of these features may affect infectious diseases. Therefore, through the above steps S310 to S340, features with strong correlation between the cumulative effects of features and the outbreak of infectious diseases can be screened out according to the cumulative values of feature values, thereby improving the accuracy of determining infectious disease warning features.

[0130] In an alternative embodiment, a third virtual score can also be determined. In this way, the specific implementation manner of step S140 can include determining the target feature for warning the second infectious disease according to the multiple first correlation degrees and the third virtual score, so as to warn the second infectious disease according to the target feature; the specific implementation manner of step S140 can also include determining the target feature for warning the second infectious disease according to the multiple first correlation degrees, the second virtual score and the third virtual score, so as to warn the second infectious disease according to the target feature.

[0131] Exemplarily, Figure 4 The flowchart shows a method for determining the third virtual score in an exemplary embodiment of the present disclosure. Refer to Figure 4 and this method may include steps S410 to S440.

[0132] In step S410, with a preset time interval as the target period, a second preset number of target periods are determined before the historical propagation duration period.

[0133] In an alternative embodiment, the duration corresponding to the preset time interval can be custom-set. When custom-setting, it can be set with reference to experience or directly randomly, that is, the duration of the target period can be custom-set.

[0134] In an alternative embodiment, the duration corresponding to the preset time interval can be less than or equal to the duration corresponding to the historical propagation duration period.

[0135] Taking the duration corresponding to the preset time interval being equal to the duration T of the historical propagation duration period as an example, starting from the start date of the historical propagation duration period, before the start date of the historical propagation duration period, a second preset number of target periods with a duration of T are determined.

[0136] Among them, the second preset number of target periods can be consecutive or non-consecutive, and the value of the second preset number can also be custom-configured, and this exemplary embodiment does not make special limitations on this. For example, if the preset time interval is 20 days and the start date of the historical propagation duration period is April 1, 2022, then the 20 days before April 1, 2022 can be determined as target period 1, the 40 days before April 1, 2022 to the 20 days before April 1, 2022 can be determined as target period 2, the 60 days before April 1, 2022 to the 40 days before April 1, 2022 can be determined as target period 3, and so on. Multiple target periods can be determined before the historical propagation duration period, and then, any second preset number of target periods can be selected from them.

[0137] In step S420, according to the preset sampling frequency, the candidate feature is sampled in each of the target periods respectively to determine the to-be-compared feature sequences corresponding to the candidate feature in each of the target periods respectively.

[0138] Taking the preset sampling frequency being once a day as an example, for each candidate feature, the feature value of the candidate feature every day can be collected in each of the second preset number of target periods to generate the feature sequence of the candidate feature in the target period, and this feature sequence is used as the to-be-compared feature sequence. In other words, the number of to-be-compared feature sequences corresponding to each candidate feature is the second preset number.

[0139] Next, in step S430, based on the preset sampling frequency, in any one of the target periods within the historical propagation duration period, the candidate feature is sampled to obtain the target feature sequence of the candidate feature.

[0140] In an alternative embodiment, the start date of the historical propagation duration is the start date of the first target period within the historical propagation duration. Taking the duration of the historical propagation duration as 60 days and the duration of the target period as 10 days as an example, the historical propagation duration includes 6 target periods. The 1st day to the 10th day within the historical propagation duration is the 1st target period within the historical propagation duration, the 11th day to the 20th day within the historical propagation duration is the 2nd target period within the historical propagation duration, and so on. The 51st day to the 60th day within the historical propagation duration is the 6th target period within the historical propagation duration.

[0141] Taking the preset sampling frequency as daily as an example, for each candidate feature, the feature value of the candidate feature can be collected every day within any target period within the historical propagation duration to obtain the target feature sequence of the candidate feature.

[0142] In step S440, according to the degree of difference between the to-be-compared feature sequence and the target feature sequence, the third virtual score of the candidate feature is determined.

[0143] Exemplarily, the degree of difference can be determined in the following manner: calculate the second correlation degree between each to-be-compared feature sequence and the target feature sequence respectively to obtain the second preset number of second correlation degrees; based on the negative correlation relationship, determine the degree of difference between the to-be-compared feature sequence and the target feature sequence according to the mean value of the second preset number of second correlation degrees.

[0144] The larger the mean value of the second correlation degrees, the smaller the degree of difference, that is, the degree of difference is negatively correlated with the second preset number of second correlation degrees. Therefore, based on the negative correlation relationship, the degree of difference between the to-be-compared feature sequence of the candidate feature and the target feature sequence can be determined according to the mean value of the second preset number of second correlation degrees, so as to characterize the periodic difference degree of the candidate feature. Among them, the second correlation degree can be determined based on any one of the Pearson correlation coefficient, T-test, and rank sum test. It can also be determined based on other methods for determining correlation, such as covariance, etc. This exemplary embodiment does not make special limitations on this.

[0145] In an alternative embodiment, the degree of difference can be determined according to the opposite of the mean value corresponding to the second degree of correlation of the second preset quantity. In this way, while satisfying that the degree of difference is negatively correlated with the mean value of the second degree of correlation, it can be ensured that the negative correlation relationship between the two is linear, which is convenient for weighted fusion with the first virtual score and / or the second virtual score. Of course, the degree of difference can also be determined according to other relationships that make the mean value of the second degree of correlation and the degree of difference negatively correlated. For example, the degree of difference is the reciprocal of the second preset quantity of the second degree of correlation. This exemplary embodiment does not make special limitations on this.

[0146] When the second preset quantity is a positive integer greater than 1, the periodic differences of the candidate features can be determined by comparing the feature sequences within multiple periods, so as to improve the accuracy of determining the degree of difference, and further improve the accuracy of the determined third virtual score.

[0147] Exemplarily, the specific implementation manner of step S440 can be to determine the third virtual score according to the positive proportional relationship between the degree of difference and the third virtual score, that is, the greater the degree of difference, the greater the third virtual score. For example, the degree of difference can be directly determined as the third virtual score.

[0148] When determining the second degree of correlation based on the Pearson correlation coefficient, since the correlation coefficient determined by the Pearson correlation coefficient can be positive or negative, but whether it is positive or negative, the greater the absolute value of the correlation coefficient, the higher the degree of correlation. Therefore, when representing the second degree of correlation based on the Pearson coefficient, the absolute value of each second degree of correlation can be taken first, and then the mean value of the absolute values is obtained. Furthermore, according to the opposite of the mean value corresponding to each absolute value, the degree of difference between the to-be-compared feature sequence and the target feature sequence is determined. In this way, the relationship that the degree of difference is negatively correlated with the mean value of the second preset quantity of the second degree of correlation can be satisfied.

[0149] For example, the periodically changing features cannot be determined by the correlation between the feature sequence within the historical propagation duration and the new number sequence within the historical propagation duration. For example, during the period of infectious disease transmission, the periodic features themselves have a large difference from the previous periods, but there is no direct correlation with the newly added number of infections. The changes in the time series within such a period are not sufficient to show an obvious difference in the cumulative value. Then, the differences within the period need to be considered. The greater the difference within the period sequence, the greater its impact on the transmission of infectious diseases.

[0150] Specifically, even if the period of the feature sequence Fi of the candidate feature i is unknown, the feature values of the candidate feature i can be intercepted multiple times at a fixed time difference, that is, with a preset time interval as the target period, to obtain multiple feature sequences corresponding to multiple target periods. For periodic features, as long as the intercepted time period is fixed, the intercepted feature sequences should also be periodic. Therefore, the differences between the feature sequences of the candidate features in different periods can be determined based on the feature sequences within each target period.

[0151] Taking the preset time interval as the duration T corresponding to the historical transmission duration period of the first infectious disease, the second preset quantity as 3, the preset sampling frequency as daily, and the second correlation degree determined based on the Pearson coefficient as an example, there is only one target period within the historical transmission duration period, that is, the entire historical transmission duration period. Then, the feature values of the candidate feature are collected daily within the historical transmission duration period. In this way, the target feature sequence F corresponding to the candidate feature within the historical transmission duration period can be obtained. iT Then, starting from the start date a of the historical transmission duration period as the starting time point, three target periods are intercepted at an interval of the duration T. The feature values of the candidate feature are sampled daily in each target period, and correspondingly, three feature sequences F to be compared are obtained. i-1T 、F i-2T 、F i-3T The third virtual score of each candidate feature i can be obtained based on the following formula (2):

[0152]

[0153] In formula (2), ZF i represents the finally obtained third virtual score, pearson represents the Pearson correlation coefficient, and abs represents taking the absolute value. The higher the Pearson correlation coefficient, the higher the correlation degree between the feature sequence to be compared and the target sequence, then the lower the difference degree, and the lower the difference degree, the smaller the third virtual score.

[0154] In an optional implementation manner, as described above, after determining the third virtual score, the target feature for warning the second infectious disease can be selected from the candidate features based on the multiple first correlation degrees corresponding to the candidate feature and the third virtual score.

[0155] Exemplarily, based on the multiple first correlation degrees and the third virtual score, selecting target features for warning of a second infectious disease from the candidate features may include: determining the maximum first correlation degree from the multiple first correlation degrees; determining the first virtual score of the candidate feature according to the maximum first correlation degree; respectively determining the first product of the first virtual score and the first weight corresponding to the first virtual score, and the third product of the third virtual score and the third weight corresponding to the second virtual score; determining the second target virtual score of the candidate feature according to the sum of the first product and the third product; and selecting target features for warning of a second infectious disease from the candidate features based on the second target virtual score.

[0156] Among them, for the specific implementation of selecting target features based on the second target virtual score, reference may also be made to the above specific implementation of selecting target features based on the first virtual score, and the technical term "first virtual score" therein may be replaced with "second target virtual score", which will not be elaborated here.

[0157] Similarly, when there are multiple historical transmission duration periods of a certain first infectious disease, the above steps may be respectively executed according to each historical transmission duration period to respectively obtain the second target virtual scores of each feature in each historical transmission duration period, and then the second target virtual scores corresponding to each historical transmission duration period are weighted and averaged, and the final second target virtual scores of each candidate feature in this first infectious disease are determined according to the weighted average second target virtual scores, so as to select target features based on the final second target virtual scores.

[0158] In an alternative embodiment, as described above, after determining the third virtual score, target features for warning of a second infectious disease may also be determined based on the multiple first correlation degrees, the second virtual score, and the third virtual score.

[0159] Exemplarily, based on the multiple first correlation degrees, second virtual scores, and third virtual scores corresponding to each candidate feature, the specific implementation of selecting target features for early warning of the second infectious disease from the candidate features may include: determining the maximum first correlation degree from the multiple first correlation degrees; determining the first virtual score of the candidate feature according to the maximum first correlation degree; respectively determining the first product of the first virtual score and the first weight corresponding to the first virtual score, the second product of the second virtual score and the second weight corresponding to the second virtual score, and the third product of the third virtual score and the third weight corresponding to the third virtual score; calculating the sum of the first product, the second product, and the third product to determine the third target virtual score corresponding to the candidate feature in the first infectious disease; and selecting target features for early warning of the second infectious disease from the candidate features based on the third target virtual score.

[0160] For example, after determining the third virtual score of each candidate feature, the first virtual score, second virtual score, and third virtual score of each candidate feature may be weighted and summed according to the first weight, second weight, and third weight to determine the third target virtual score of each candidate feature, and then target features may be selected from the candidate features according to the determined third target virtual score of each candidate feature.

[0161] Exemplarily, when the first infectious disease is the same as the second infectious disease, the candidate features may be sorted in descending order based on the third target virtual score corresponding to the candidate features; and the target features for early warning of the first infectious disease may be determined according to the first M candidate features in the sorting result, where M is a positive integer.

[0162] Exemplarily, when the first infectious disease is different from the second infectious disease, the above steps may be respectively executed for multiple different first infectious diseases to obtain the third target virtual scores of the candidate features in multiple different first infectious diseases, and then the target features for early warning of the second infectious disease may be determined. Wherein, the first infectious disease is a known infectious disease that has occurred, and the second infectious disease is an unknown infectious disease that has not occurred, that is, a newly emerging infectious disease.

[0163] Exemplarily, Figure 5 The flowchart showing the method for early warning of newly emerging infectious diseases in an exemplary embodiment of the present disclosure is referred to. Figure 5, the method may include steps S510 to S530. Among them: in step S510, obtain the mean value of the third target virtual scores corresponding to the candidate features in multiple different first infectious diseases; in step S520, sort the candidate features in descending order according to the mean value of the third target virtual scores; in step S530, determine the target features for warning the second infectious disease according to the top M candidate features in the sorting result, where M is a positive integer.

[0164] For example, relevant data of multiple different first infectious diseases can be obtained, and then for each first infectious disease, the above steps can be executed to obtain the third target virtual scores of each candidate feature in this first infectious disease. Then, calculate the mean value of the third target virtual scores of each candidate feature in multiple first infectious diseases, sort each candidate feature according to the descending order of this mean value, select the top M candidate features in the sorting result, and use them as the target features for warning the second infectious disease.

[0165] Similarly, when the historical transmission duration of a certain first infectious disease includes multiple periods, the above steps can be executed respectively according to each historical transmission duration to obtain the third target virtual scores corresponding to each feature in each historical transmission duration respectively. Then, perform weighted averaging on the third target virtual scores corresponding to each historical transmission duration, and determine the final third target virtual score of each candidate feature in this first infectious disease based on the weighted average third target virtual score, so as to select the target features based on the final third target virtual score.

[0166] In an alternative embodiment, the above first weight, second weight, and third weight can be set customarily. Among them, the first weight can be greater than the sum of the first weight and the second weight. That is, the first weight can be the maximum of the three weights. This is because the first virtual score determined directly based on the correlation between the feature sequence and the new case sequence has a relatively high confidence level, so the corresponding first weight is the largest. For example, the first weight can be set to 1, and both the second weight and the third weight can be set to 0.5. In practice, the weights of the first virtual score, the second virtual score, and the third virtual score can be adjusted customarily according to the actual scenario.

[0167] In an alternative embodiment, since the first virtual score has a relatively high confidence level, the second weight and the third weight can be determined respectively according to the error between the second virtual score and the first virtual score, and the error between the third virtual score and the first virtual score.

[0168] Exemplarily, Figure 6 shows a schematic flowchart of a method for determining the second weight and the third weight in an exemplary embodiment of the present disclosure. Refer toFigure 6 , the method may include steps S610 to S630. Among them:

[0169] In step S610, the candidate features are sorted according to the descending order of the maximum first correlation degree corresponding to the candidate features, and the first N reference candidate features are determined.

[0170] Wherein, N is a positive integer, which can be customized according to requirements. For example, the candidate features with the first 20 largest first correlation degrees in the sorting can be selected as the reference candidate features. Since the maximum first correlation degree is proportional to the first virtual score, the first N candidate features can also be sorted in descending order according to the first virtual score to determine the reference candidate features.

[0171] In step S620, for each reference candidate feature, the first error between the maximum first correlation degree of the reference candidate feature and the second virtual score of the reference candidate feature, and the second error between the maximum first correlation degree of the reference candidate feature and the third virtual score of the reference candidate feature are calculated respectively.

[0172] In an alternative embodiment, the first error may also be the error between the first virtual score and the second virtual score corresponding to the reference candidate feature, and the second error may also be the error between the first virtual score and the third virtual score corresponding to the reference candidate feature.

[0173] Among them, the first error and the second error can be characterized by the Root Mean Square Error (RMSE), or can be characterized by other errors, such as the Mean Absolute Error (MAE), the cross-entropy loss function, etc. This exemplary embodiment does not make special limitations on this.

[0174] Next, in step S630, according to the first error and the second error corresponding to each reference candidate feature, the second weight and the third weight are determined.

[0175] Exemplarily, the specific implementation manner of step S630 may include: calculating the first error mean of the first errors corresponding to the N reference candidate features and the second error mean of the second errors corresponding to the N reference candidate features respectively; generating an input vector according to the opposite number of the first error mean and the opposite number of the second error mean, and inputting the input vector into a probability mapping function to obtain a first probability corresponding to the second virtual score and a second probability corresponding to the third virtual score; determining the second weight corresponding to the second virtual score according to the first probability, and determining the third weight corresponding to the third virtual score according to the second probability.

[0176] In an alternative embodiment, the probability mapping function may include any function capable of mapping an input vector to a probability distribution, such as the softmax function, the sigmoid function, etc. This exemplary embodiment does not make any special limitations thereon.

[0177] Next, taking the selected reference candidate features being 20, the probability mapping function being the softmax function, and the first error and the second error being characterized by the root mean square error as an example, a specific implementation manner of determining the first weight and the second weight based on the first error and the second error will be further described.

[0178] For the selected N reference candidate features, the corresponding second virtual score and third virtual score can be calculated. Then, calculate the first error between the second virtual score of each reference candidate feature and the first virtual score, and the second error between the third virtual score and the first virtual score. Then, respectively, through the following formulas (3) and (4), obtain the mean value of the first error and the mean value of the second error:

[0179]

[0180]

[0181] In formulas (3) and (4), represents the first virtual score or the maximum first correlation degree of the reference candidate feature i, is the second virtual score of feature i, is the third virtual score of feature i, ε(C(F s )) is the mean value of the first error, and ε(Z(F s )) is the mean value of the second error.

[0182] Taking the opposite numbers of the mean value of the first error and the mean value of the second error obtained based on formulas (3) and (4) as the input vector, and inputting it into the softmax function, the softmax function can obtain the second weight and the third weight respectively through the following formulas (5) and (6):

[0183]

[0184]

[0185] In formulas (5) and (6), w c is the second weight, w zis the third weight. It can be seen from formulas (5) and (6) that the sum of the two probability values obtained by mapping through the softmax function is 1, and the probability values obtained by mapping are used as the corresponding weights. That is, the probability value mapped according to the opposite of the mean of the first error is the second weight, and the probability value mapped according to the opposite of the mean of the second error is the third weight. That is to say, the sum of the second weight and the third weight is 1. Among them, the greater the error, the smaller the probability value (i.e., weight) obtained after the softmax transformation.

[0186] In the present disclosure, the first virtual score is determined based on the idea of scanning phase difference. By obtaining multiple feature sequences corresponding to multiple candidate time periods through scanning phase difference, features that have a lagging effect on the spread of infectious diseases can be screened based on the correlation degree between the multiple feature sequences and the daily new number sequence, so as to improve the accuracy of determining the warning features of infectious diseases.

[0187] Furthermore, the second virtual score in the present disclosure can be understood as the cumulative score of candidate features, which can measure the impact of the cumulative effect of candidate features on the spread of infectious diseases. The third virtual score can be understood as the periodic score, which can measure the impact of the periodic difference of features on the spread of infectious diseases. By selecting target features through the second virtual score and the third virtual score, features with a large correlation degree between the cumulative effect and the daily new number of infectious diseases, and features with a large correlation degree between the periodic difference and the daily new number of infectious diseases can be further screened, thereby further improving the accuracy of determining the warning features of infectious diseases, and thus more timely and accurate warnings of infectious diseases can be made based on more accurate warning features.

[0188] Those skilled in the art can understand that all or part of the steps of implementing the above embodiments are realized by a computer program executed by a CPU. When the computer program is executed by the CPU, the above functions defined by the above method provided by the present invention are executed. The program can be stored in a computer-readable storage medium, and the storage medium can be a read-only memory, a magnetic disk or an optical disc, etc.

[0189] In addition, it should be noted that the above drawings are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present invention, rather than for limiting purposes. It is easy to understand that the processes shown in the above drawings do not indicate or limit the time sequence of these processes. Additionally, it is also easy to understand that these processes can be executed synchronously or asynchronously in, for example, multiple modules.

[0190] Figure 7 Shows a schematic structural diagram of an infectious disease warning device in an exemplary embodiment of the present disclosure. Refer to Figure 7, the device 700 may include a new case number sequence determination module 710, a plurality of feature sequence acquisition modules 720, a first correlation degree calculation module 730, and an infectious disease early warning module 740. Among them: The new case number sequence determination module 710 is configured to obtain the new case number sequence corresponding to the first infectious disease determined based on a preset sampling frequency during the historical transmission duration of the first infectious disease; The plurality of feature sequence acquisition modules 720 are configured to obtain a plurality of feature sequences corresponding to candidate features in a plurality of candidate time periods, the duration of each candidate time period being the same as the duration corresponding to the historical transmission duration, and the time difference between the start date of each candidate time period and the start date of the historical transmission duration being within a first preset value, and the feature sequence corresponding to each candidate time period being determined according to the feature values collected based on the preset sampling frequency during the candidate time period; The first correlation degree calculation module 730 is configured to calculate the first correlation degree between the feature sequence corresponding to the candidate feature in each candidate time period and the new case number sequence respectively, so as to obtain a plurality of first correlation degrees corresponding to the plurality of feature sequences; The infectious disease early warning module 740 is configured to select, according to the plurality of first correlation degrees, a target feature for warning the second infectious disease from the candidate features, so as to warn the second infectious disease according to the target feature.

[0191] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the infectious disease early warning module 740 may be specifically configured to: determine the maximum first correlation degree from the plurality of first correlation degrees corresponding to the plurality of feature sequences of the candidate features; determine the first virtual score of the candidate feature according to the maximum first correlation degree corresponding to the candidate feature; select, according to the first virtual score, a target feature for warning the second infectious disease from the candidate features.

[0192] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the infectious disease early warning module 740 may further be specifically configured to: obtain a first cumulative value of the feature values of the candidate features collected based on the preset sampling frequency within a first target time period, where the end date of the first target time period is the start date of the historical transmission duration or a date before the start date of the historical transmission duration, and the duration of the first target time period is a second preset value; determine a first preset number of historical same - period time periods corresponding to the first target time period, and respectively collect the feature values of the candidate features based on the preset sampling frequency within each historical same - period time period; respectively determine second cumulative values of the feature values of the candidate features collected within each historical same - period time period, to obtain the first preset number of second cumulative values; determine a second virtual score of the candidate feature according to the first cumulative value and the first preset number of second cumulative values; and select a target feature for early warning of a second infectious disease from the candidate features based on the plurality of first correlation degrees and the second virtual score.

[0193] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the determining a second virtual score of the candidate feature according to the first cumulative value and the first preset number of second cumulative values includes: determining an average value of the first preset number of second cumulative values; determining an absolute value of the difference between the first cumulative value and the average value; and calculating a ratio between the absolute value and the average value to determine the second virtual score of the candidate feature.

[0194] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the selecting a target feature for early warning of a second infectious disease from the candidate features based on the plurality of first correlation degrees and the second virtual score includes: determining the maximum first correlation degree corresponding to the candidate feature from the plurality of first correlation degrees corresponding to the plurality of feature sequences of the candidate feature; determining a first virtual score of the candidate feature according to the maximum first correlation degree corresponding to the candidate feature; respectively determining a first product of the first virtual score and a first weight corresponding to the first virtual score, and a second product of the second virtual score and a second weight corresponding to the second virtual score; determining a first target virtual score of the candidate feature according to the sum of the first product and the second product; and selecting a target feature for early warning of a second infectious disease from the candidate features based on the first target virtual score.

[0195] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, selecting, from the candidate features, target features for warning of a second infectious disease based on the plurality of first correlation degrees and the second virtual score includes: taking a preset time interval as a target period, and determining a second preset number of target periods before the historical transmission duration period; sampling the candidate features in each of the target periods according to the preset sampling frequency to respectively determine corresponding to-be-compared feature sequences of the candidate features in each of the target periods; sampling the candidate features in any one of the target periods within the historical transmission duration period based on the preset sampling frequency to obtain a target feature sequence of the candidate features; determining a third virtual score of the candidate features according to the difference degree between the to-be-compared feature sequences and the target feature sequence; and selecting, from the candidate features, target features for warning of a second infectious disease based on the plurality of first correlation degrees, the second virtual score, and the third virtual score.

[0196] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the infectious disease warning module 740 may further be configured to: take a preset time interval as a target period, and determine a second preset number of target periods before the historical transmission duration period; sample the candidate features in each of the target periods according to the preset sampling frequency to respectively determine corresponding to-be-compared feature sequences of the candidate features in each of the target periods; sample the candidate features in any one of the target periods within the historical transmission duration period based on the preset sampling frequency to obtain a target feature sequence of the candidate features; determine a third virtual score of the candidate features according to the difference degree between the to-be-compared feature sequences and the target feature sequence; and select, from the candidate features, target features for warning of a second infectious disease based on the plurality of first correlation degrees and the third virtual score.

[0197] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, selecting, from the candidate features, target features for warning of a second infectious disease based on the plurality of first correlation degrees and the third virtual score includes: determining the maximum first correlation degree corresponding to the candidate feature from among the plurality of first correlation degrees corresponding to the plurality of feature sequences of the candidate feature; determining a first virtual score of the candidate feature according to the maximum first correlation degree corresponding to the candidate feature; respectively determining a first product of the first virtual score and a first weight corresponding to the first virtual score, and a third product of the third virtual score and a third weight corresponding to the second virtual score; determining a second target virtual score of the candidate feature according to the sum of the first product and the third product; and selecting, based on the second target virtual score, target features for warning of a second infectious disease from the candidate features.

[0198] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the degree of difference is determined by the following method: respectively calculating a second correlation degree between each of the to-be-compared feature sequences and the target feature sequence to obtain the second preset number of second correlation degrees; and determining the degree of difference between the to-be-compared feature sequence and the target feature sequence according to the mean value of the second preset number of second correlation degrees, where the degree of difference is negatively correlated with the mean value of the second correlation degrees.

[0199] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, selecting, from the candidate features, target features for warning of a second infectious disease based on the plurality of first correlation degrees, the second virtual score, and the third virtual score includes: determining the maximum first correlation degree corresponding to the candidate feature from among the plurality of first correlation degrees corresponding to the plurality of feature sequences of the candidate feature; determining a first virtual score of the candidate feature according to the maximum first correlation degree corresponding to the candidate feature; respectively determining a first product of the first virtual score and a first weight corresponding to the first virtual score, a second product of the second virtual score and a second weight corresponding to the second virtual score, and a third product of the third virtual score and a third weight corresponding to the third virtual score; calculating the sum of the first product, the second product, and the third product to determine a third target virtual score of the candidate feature in the first infectious disease; and selecting, based on the third target virtual score, target features for warning of a second infectious disease from the candidate features.

[0200] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the second weight and the third weight are determined as follows: The candidate features are sorted according to the descending order of the maximum first correlation degree corresponding to the candidate features, and the first N reference candidate features are determined, where N is a positive integer; for each reference candidate feature, the first error between the maximum first correlation degree of the reference candidate feature and the second virtual score of the reference candidate feature, and the second error between the maximum first correlation degree of the reference candidate feature and the third virtual score of the reference candidate feature are calculated respectively; according to the first error and the second error corresponding to each reference candidate feature, the second weight and the third weight are determined.

[0201] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the determining the second weight and the third weight according to the first error and the second error corresponding to each reference candidate feature includes: calculating the first error mean of the first errors corresponding to the N reference candidate features and the second error mean of the second errors corresponding to the N reference candidate features respectively; generating an input vector according to the opposite number of the first difference mean and the opposite number of the second error mean, and inputting the input vector into a probability mapping function to obtain a first probability corresponding to the second virtual score and a second probability corresponding to the third virtual score; determining the second weight corresponding to the second virtual score according to the first probability, and determining the third weight corresponding to the third virtual score according to the second probability.

[0202] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the first infectious disease and the second infectious disease are the same; the selecting the target features for warning the second infectious disease from the candidate features based on the third target virtual score includes: sorting the candidate features in descending order based on the third target virtual score; determining the target features for warning the first infectious disease according to the first M candidate features in the sorting result, where M is a positive integer.

[0203] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the first infectious disease and the second infectious disease are different; the selecting the target features for warning the second infectious disease from the candidate features based on the third target virtual score includes: obtaining the mean value of the third target virtual scores corresponding to the candidate features in multiple different first infectious diseases; sorting the candidate features in descending order according to the mean value of the third target virtual scores; determining the target features for warning the second infectious disease according to the first M candidate features in the sorting result, where M is a positive integer.

[0204] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the first degree of correlation is determined based on any one of the Pearson correlation coefficient, the T-test, and the rank sum test.

[0205] The specific details of each unit in the above-mentioned infectious disease warning device have been described in detail in the corresponding infectious disease warning method, and thus will not be elaborated here.

[0206] It should be noted that although several modules or units of a device for action execution are mentioned in the foregoing detailed description, such a division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more of the above-mentioned modules or units can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0207] In addition, although the steps of the methods in the present disclosure are described in a specific order in the drawings, this does not require or imply that these steps must be executed in that specific order, or that all the steps shown must be executed to achieve the desired result. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step for execution, and / or one step may be decomposed into multiple steps for execution, etc.

[0208] Through the description of the above embodiments, those skilled in the art can easily understand that the exemplary embodiments described herein can be implemented by software, or by a combination of software and necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a mobile terminal, or a network device, etc.) to execute the methods according to the embodiments of the present disclosure.

[0209] In an exemplary embodiment of the present disclosure, a computer storage medium capable of implementing the above method is further provided. A program product capable of implementing the methods described in this specification is stored thereon. In some possible embodiments, various aspects of the present disclosure can also be implemented in the form of a program product, which includes program code. When the program product runs on a terminal device, the program code is used to cause the terminal device to execute the steps according to various exemplary embodiments of the present disclosure described in the "Exemplary Methods" section of this specification.

[0210] A program product for implementing the above method according to an embodiment of the present disclosure may be a portable compact disc read-only memory (CD-ROM), include program code, and may run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto. In this document, a readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0211] The program product may adopt any combination of one or more readable media. The readable media may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0212] The computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries the readable program code. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable signal medium may also be any readable medium other than the readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0213] The program code contained on the readable medium may be transmitted by any appropriate medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination of the above.

[0214] Program code for performing the operations of the present disclosure may be written in any combination of one or more programming languages, including object-oriented programming languages such as Java, C++, etc., and also including conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computing device, partially on the user's device, executed as a stand-alone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving a remote computing device, the remote computing device may be connected to the user's computing device through any kind of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0215] In addition, in an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method is also provided.

[0216] Those skilled in the art can understand that various aspects of the present disclosure can be implemented as a system, method, or program product. Therefore, various aspects of the present disclosure can be specifically implemented in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or an implementation combining hardware and software aspects, which can be collectively referred to as "circuitry", "module", or "system" here.

[0217] The following refers to Figure 8 to describe the electronic device 800 according to such an embodiment of the present disclosure. Figure 8 The electronic device 800 shown is only an example and should not impose any limitations on the functions and usage scope of the embodiments of the present disclosure.

[0218] As Figure 8 shown, the electronic device 800 is presented in the form of a general-purpose computing device. The components of the electronic device 800 may include, but are not limited to: at least one of the above-mentioned processing units 810, at least one of the above-mentioned storage units 820, a bus 830 connecting different system components (including the storage unit 820 and the processing unit 810), and a display unit 840.

[0219] Among them, the storage unit stores program code, and the program code can be executed by the processing unit 810, so that the processing unit 810 executes the steps according to various exemplary embodiments of the present disclosure described in the above "Exemplary Method" section of this specification. For example, the processing unit 810 may execute the various steps as Figures 1 to 6 shown in

[0220] The storage unit 820 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 8201 and / or a cache storage unit 8202, and may further include a read-only storage unit (ROM) 8203.

[0221] The storage unit 820 may also include a program / utilities 8204 having a set (at least one) of program modules 8205. Such program modules 8205 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment.

[0222] The bus 830 may represent one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus structures.

[0223] The electronic device 800 may also communicate with one or more external devices 900 (such as a keyboard, a pointing device, a Bluetooth device, etc.), and may also communicate with one or more devices that enable a user to interact with the electronic device 800, and / or communicate with any device that enables the electronic device 800 to communicate with one or more other computing devices (such as a router, a modem, etc.). Such communication may be through an input / output (I / O) interface 850. Further, the electronic device 800 may also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 860. As shown in the figure, the network adapter 860 communicates with other modules of the electronic device 800 through the bus 830. It should be understood that although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 800, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0224] Through the description of the above embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or by a combination of software and necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the embodiments of the present disclosure.

[0225] In addition, the above-mentioned drawings are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present disclosure, rather than for limiting purposes. It is easy to understand that the processes shown in the above-mentioned drawings do not indicate or limit the chronological order of these processes. Additionally, it is also easy to understand that these processes can be executed, for example, synchronously or asynchronously in multiple modules.

[0226] Other embodiments of the present disclosure will be readily envisioned by those of ordinary skill in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include known common knowledge or conventional technical means in the technical field not disclosed by the present disclosure. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of the present disclosure are pointed out by the claims.

Claims

1. An infectious disease early warning method, characterized in that, it includes: Obtain the sequence of the number of newly added cases corresponding to the first infectious disease determined based on a preset sampling frequency during the historical transmission duration of the first infectious disease; Obtain multiple feature sequences corresponding to a candidate feature in multiple candidate time periods. The duration of each candidate time period is the same as the duration of the historical transmission duration, and the time difference between the start date of each candidate time period and the start date of the historical transmission duration is within a first preset value. The feature sequence corresponding to each candidate time period is determined according to the feature values collected based on the preset sampling frequency during the candidate time period; Calculate the first correlation degree between the feature sequence corresponding to the candidate feature in each candidate time period and the sequence of the number of newly added cases respectively, so as to obtain multiple first correlation degrees corresponding to multiple feature sequences; According to the multiple first correlation degrees, select a target feature for early warning of the second infectious disease from the candidate features, so as to early warn the second infectious disease according to the target feature; The step of selecting a target feature for early warning of the second infectious disease from the candidate features according to the multiple first correlation degrees includes: obtaining the first cumulative value of the feature values of the candidate feature collected based on the preset sampling frequency during a first target time period. The end date of the first target time period is the start date of the historical transmission duration or a date before the start date of the historical transmission duration, and the duration of the first target time period is a second preset value; determining the first preset number of historical synchronous time periods corresponding to the first target time period, and respectively collecting the feature values of the candidate feature based on the preset sampling frequency in each historical synchronous time period; respectively determining the second cumulative value of the feature values of the candidate feature collected in each historical synchronous time period to obtain the first preset number of second cumulative values; determining the second virtual score of the candidate feature according to the first cumulative value and the first preset number of second cumulative values; based on the multiple first correlation degrees and the second virtual score, select a target feature for early warning of the second infectious disease from the candidate features; when the first infectious disease and the second infectious disease are different, the step of selecting a target feature for early warning of the second infectious disease from the candidate features based on the multiple first correlation degrees and the second virtual score includes: for each candidate feature, determining the first target virtual score of the candidate feature in multiple first infectious diseases based on the multiple first correlation degrees and the second virtual score; determining the mean value of the first target virtual scores; sorting the candidate features in descending order according to the mean value, and determining the first M candidate features in the sorting result as the target features for early warning of the second infectious disease; or For each of the multiple first infectious diseases, based on the multiple first degrees of correlation and the second virtual score, select the target feature corresponding to the first infectious disease from the candidate features, and determine the target feature for warning of the second infectious disease according to the intersection of the target features corresponding to the multiple first infectious diseases.

2. The infectious disease warning method according to claim 1, wherein, the step of selecting the target feature for warning of the second infectious disease from the candidate features according to the multiple first degrees of correlation includes: determine the maximum first degree of correlation corresponding to the candidate feature from the multiple first degrees of correlation corresponding to the multiple feature sequences of the candidate feature; determine the first virtual score of the candidate feature according to the maximum first degree of correlation corresponding to the candidate feature; select the target feature for warning of the second infectious disease from the candidate features according to the first virtual score.

3. The infectious disease warning method according to claim 1, wherein, the step of determining the second virtual score of the candidate feature according to the first cumulative value and the first preset number of second cumulative values includes: determine the average value of the first preset number of second cumulative values; determine the absolute value of the difference between the first cumulative value and the average value; calculate the ratio between the absolute value and the average value to determine the second virtual score of the candidate feature.

4. The infectious disease warning method according to claim 1, wherein, the step of selecting the target feature for warning of the second infectious disease from the candidate features based on the multiple first degrees of correlation and the second virtual score includes: determine the maximum first degree of correlation corresponding to the candidate feature from the multiple first degrees of correlation corresponding to the multiple feature sequences of the candidate feature; determine the first virtual score of the candidate feature according to the maximum first degree of correlation corresponding to the candidate feature; respectively determine the first product of the first virtual score and the first weight corresponding to the first virtual score, and the second product of the second virtual score and the second weight corresponding to the second virtual score; determine the first target virtual score of the candidate feature according to the sum of the first product and the second product; select the target feature for warning of the second infectious disease from the candidate features based on the first target virtual score.

5. The infectious disease warning method according to claim 1, wherein, the step of selecting the target feature for warning of the second infectious disease from the candidate features according to the multiple first degrees of correlation includes: taking a preset time interval as the target period, and determining a second preset number of target periods before the historical propagation duration period; sampling the candidate feature at each target period according to the preset sampling frequency to respectively determine the to-be-compared feature sequences corresponding to the candidate feature at each target period; Based on the preset sampling frequency, sample the candidate features within any one of the target periods during the historical propagation duration to obtain the target feature sequence of the candidate features; Determine the third virtual score of the candidate feature according to the degree of difference between the feature sequence to be compared and the target feature sequence; Based on the multiple first correlation degrees and the third virtual score, select the target features for warning against the second infectious disease from the candidate features.

6. The infectious disease warning method according to claim 1, wherein, the selecting the target features for warning against the second infectious disease from the candidate features based on the multiple first correlation degrees and the second virtual score includes: Taking a preset time interval as the target period, determine a second preset number of target periods before the historical propagation duration; Sample the candidate features within each of the target periods according to the preset sampling frequency to respectively determine the feature sequences to be compared corresponding to the candidate features within each of the target periods; Based on the preset sampling frequency, sample the candidate features within any one of the target periods during the historical propagation duration to obtain the target feature sequence of the candidate features; Determine the third virtual score of the candidate feature according to the degree of difference between the feature sequence to be compared and the target feature sequence; Based on the multiple first correlation degrees, the second virtual score and the third virtual score, select the target features for warning against the second infectious disease from the candidate features.

7. The infectious disease warning method according to claim 5 or 6, wherein, the degree of difference is determined by the following method: Calculate the second correlation degree between each of the feature sequences to be compared and the target feature sequence respectively to obtain the second preset number of second correlation degrees; Based on the negative correlation relationship, determine the degree of difference between the feature sequence to be compared and the target feature sequence according to the mean value of the second preset number of second correlation degrees.

8. The infectious disease warning method according to claim 5, wherein, the selecting the target features for warning against the second infectious disease from the candidate features based on the multiple first correlation degrees and the third virtual score includes: Determine the maximum first correlation degree corresponding to the candidate feature from the multiple first correlation degrees corresponding to the multiple feature sequences of the candidate feature; According to the maximum first correlation degree corresponding to the candidate feature, determine the first virtual score of the candidate feature; respectively determine the first product of the first virtual score and the first weight corresponding to the first virtual score, and the third product of the third virtual score and the third weight corresponding to the second virtual score; Determine the second target virtual score of the candidate feature according to the sum of the first product and the third product; Based on the second target virtual score, select the target features for warning against the second infectious disease from the candidate features.

9. The infectious disease early warning method according to claim 6, wherein, selecting, from the candidate features, target features for early warning of a second infectious disease based on the multiple first correlation degrees, the second virtual score, and the third virtual score includes: determining, from the multiple first correlation degrees corresponding to the multiple feature sequences of the candidate features, the maximum first correlation degree corresponding to the candidate features; determining a first virtual score of the candidate features according to the maximum first correlation degree corresponding to the candidate features; respectively determining a first product of the first virtual score and a first weight corresponding to the first virtual score, a second product of the second virtual score and a second weight corresponding to the second virtual score, and a third product of the third virtual score and a third weight corresponding to the third virtual score; calculating the sum of the first product, the second product, and the third product to determine a third target virtual score corresponding to the candidate features in the first infectious disease; selecting, based on the third target virtual score, target features for early warning of a second infectious disease from the candidate features.

10. The infectious disease early warning method according to claim 9, wherein, the second weight and the third weight are determined by the following method: sorting the candidate features according to the descending order of the maximum first correlation degree corresponding to the candidate features to determine the first N reference candidate features, where N is a positive integer; for each reference candidate feature, respectively calculating a first error between the maximum first correlation degree of the reference candidate feature and the second virtual score of the reference candidate feature, and a second error between the maximum first correlation degree of the reference candidate feature and the third virtual score of the reference candidate feature; determining the second weight and the third weight according to the first error and the second error corresponding to each reference candidate feature.

11. The infectious disease early warning method according to claim 10, wherein, determining the second weight and the third weight according to the first error and the second error corresponding to each reference candidate feature includes: respectively calculating a first error mean of the first errors corresponding to the N reference candidate features and a second error mean of the second errors corresponding to the N reference candidate features; generating an input vector according to the opposite number of the first error mean and the opposite number of the second error mean, and inputting the input vector into a probability mapping function to obtain a first probability corresponding to the second virtual score and a second probability corresponding to the third virtual score; determining the second weight corresponding to the second virtual score according to the first probability, and determining the third weight corresponding to the third virtual score according to the second probability.

12. The infectious disease early warning method according to claim 9, wherein, the first infectious disease and the second infectious disease are the same; selecting, based on the third target virtual score, target features for early warning of a second infectious disease from the candidate features includes: Perform a descending order sorting on the candidate features based on the third target virtual score; Determine, according to the top M candidate features in the sorting result, the target features for warning of the first infectious disease, where M is a positive integer.

13. The infectious disease warning method according to claim 9, characterized in that the first infectious disease is different from the second infectious disease; The selecting, based on the third target virtual score, of the target features for warning of the second infectious disease from the candidate features includes: obtaining the mean value of the third target virtual scores corresponding to the candidate features in multiple different first infectious diseases; performing a descending order sorting on the candidate features according to the mean value of the third target virtual scores; determining, according to the top M candidate features in the sorting result, the target features for warning of the second infectious disease, where M is a positive integer.

14. The infectious disease warning method according to claim 1, characterized in that the first correlation degree is determined based on any one of the Pearson correlation coefficient, the T-test, and the rank sum test.

15. An infectious disease warning device, characterized in that it includes: a new case number sequence determination module configured to obtain the new case number sequence corresponding to the first infectious disease determined based on a preset sampling frequency during the historical transmission duration of the first infectious disease; a plurality of feature sequence acquisition modules configured to obtain a plurality of feature sequences corresponding to the candidate features in a plurality of candidate time periods, the duration of each candidate time period being the same as the duration of the historical transmission duration, and the time difference between the start date of each candidate time period and the start date of the historical transmission duration being within a first preset value, and the feature sequence corresponding to each candidate time period being determined according to the feature values collected based on the preset sampling frequency during the candidate time period; a first correlation degree calculation module configured to calculate respectively the first correlation degree between the feature sequence corresponding to each candidate feature in each candidate time period and the new case number sequence, so as to obtain a plurality of first correlation degrees corresponding to the plurality of feature sequences; an infectious disease warning module configured to select, according to the plurality of first correlation degrees, the target features for warning of the second infectious disease from the candidate features, so as to warn of the second infectious disease according to the target features; Selecting, according to the multiple first correlation degrees, target features for warning of a second infectious disease from the candidate features includes: obtaining a first cumulative value of the feature values of the candidate features collected based on the preset sampling frequency within a first target time period, where an end date of the first target time period is a start date of the historical transmission duration or a date before the start date of the historical transmission duration, and a duration of the first target time period is a second preset value; determining a first preset number of historical concurrent time periods corresponding to the first target time period, and respectively collecting the feature values of the candidate features based on the preset sampling frequency within each historical concurrent time period; respectively determining second cumulative values of the feature values of the candidate features collected within each historical concurrent time period, to obtain the first preset number of second cumulative values; determining a second virtual score of the candidate features according to the first cumulative value and the first preset number of second cumulative values; selecting, based on the multiple first correlation degrees and the second virtual score, target features for warning of a second infectious disease from the candidate features; in a case where the first infectious disease is different from the second infectious disease, the selecting, based on the multiple first correlation degrees and the second virtual score, target features for warning of a second infectious disease from the candidate features includes: for each candidate feature, determining a first target virtual score of the candidate feature in multiple first infectious diseases based on the multiple first correlation degrees and the second virtual score, determining an average value of the first target virtual scores, sorting the candidate features according to a descending order of the average value, and determining the first M candidate features in the sorting result as target features for warning of a second infectious disease; or For each of multiple first infectious diseases, selecting, based on the multiple first correlation degrees and the second virtual score, target features corresponding to the first infectious disease from the candidate features, and determining, according to an intersection of the target features corresponding to the multiple first infectious diseases, target features for warning of a second infectious disease.

16. A computer-readable medium, on which a computer program is stored, characterized in that, when the program is executed by a processor, it implements the infectious disease warning method according to any one of claims 1 to 14.

17. An electronic device, characterized in that, comprising: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the infectious disease warning method according to any one of claims 1 to 14.

Images