Health risk prediction method and equipment based on environmental data fusion, and storage medium

Abstract

The invention discloses a health risk prediction method and device based on environmental data fusion and a storage medium, and belongs to the technical field of big data, the method comprises the following steps: acquiring environmental data, the environmental data comprising a plurality of different types of data, each type of environmental data comprising a plurality of items of data taking acquisition time as a timestamp, defining the type of the environment data as a target type; preprocessing the environmental data according to the target types of the environmental data, aligning the timestamps of the environmental data of all the target types based on linear interpolation, and normalizing the environmental data after the timestamps are aligned to obtain normalized environmental data; risk parameters are determined according to the normalized environment data, risk levels are determined and output according to the risk parameters, and the risk parameters represent the probability of occurrence of high risks. The invention aims to realize fusion of different types of environmental data, so that health risk early warning influencing the health of residents is made by integrating multi-dimensional environmental data, and the body health of the residents is improved.

Description

Technical Field

[0001] The present invention relates to the field of big data technology, and in particular to a health risk prediction method, device and storage medium based on environmental data fusion. Background Art

[0002] With the advancement of urbanization, the issues of residents' health and living environment quality have gradually become the focus of public attention. There are many sources of pollution that can affect the physical health of residents, including air pollution, water pollution, etc. However, current pollution monitoring generally only targets a single field, such as testing air quality to issue air pollution warnings, or testing water quality to issue water pollution warnings, etc. While paying attention to one type of environmental pollution, it is easy to ignore the possible impacts of other environmental factors, resulting in neglecting one thing while focusing on another, making it impossible for residents to receive comprehensive health risk warnings, affecting their physical health. Summary of the Invention

[0003] The present invention provides a health risk prediction method, device and storage medium based on environmental data fusion to solve one or more technical problems existing in the prior art and at least provide a beneficial option or create conditions.

[0004] The present invention provides a health risk prediction method based on environmental data fusion, the health risk prediction method based on environmental data fusion comprising: Collecting environmental data, the environmental data including multiple different types of data, each type of environmental data including multiple data items with a collection time as a timestamp, and defining the type of the environmental data as a target type; preprocessing the environmental data according to the target type of the environmental data, aligning the timestamps of the environmental data of all the target types based on linear interpolation, and normalizing the environmental data after the timestamps are aligned to obtain normalized environmental data; A risk parameter is determined according to the normalized environmental data, and a risk level is determined and output according to the risk parameter, wherein the risk parameter represents a probability of occurrence of a high risk.

[0005] Optionally, the step of determining risk parameters according to the normalized environmental data includes: Obtaining weight parameters of the target type; For the normalized environmental data with the same timestamp, a weighted average value is calculated according to the weight parameter and the normalized environmental data, and the weighted average value is defined as a sub-parameter of the risk parameter.

[0006] Optionally, the step of determining risk parameters according to the normalized environmental data includes: Acquire event state information and conditional probability information matching the target type, and form a Bayesian network of the target type based on the event state information and the conditional probability information; Based on the Bayesian network, a Bayesian probability of high risk is determined according to the normalized environmental data, and the Bayesian probability is defined as a sub-parameter of the risk parameter.

[0007] Optionally, the step of determining risk parameters according to the normalized environmental data includes: Obtaining a logistic regression prediction model for the target type; The normalized environmental data is input into the logistic regression prediction model, the model prediction probability is determined according to the output result of the logistic regression prediction model, and the model prediction probability is defined as a sub-parameter of the risk parameter.

[0008] Optionally, the health risk prediction method based on environmental data fusion further includes: Acquire training environment data matching the target type; A preset model is trained according to the training environment data, and the parameters of the preset model are adjusted based on minimizing the logarithmic loss function until the prediction accuracy of the preset model for the training environment data is greater than or equal to the preset accuracy, thereby obtaining the logistic regression prediction model, wherein the parameters of the preset model include an intercept and a regression coefficient.

[0009] Optionally, the step of determining and outputting the risk level according to the risk parameter includes: When the risk parameter includes a sub-parameter, determining the risk level according to the sub-parameter, and outputting a health risk warning corresponding to the risk level; When the risk parameter includes more than one sub-parameter, the risk level is determined by voting based on all the sub-parameters, and a health risk warning corresponding to the risk level is output.

[0010] Optionally, the step of determining the risk level according to the sub-parameters includes: When the sub-parameter is greater than or equal to a preset probability, the risk level is determined to be high risk; When the sub-parameter is less than the preset probability, the risk level is determined to be low risk.

[0011] Optionally, before the step of normalizing the environmental data after aligning the timestamps, the method further includes: For the environmental data of each target type, noise reduction processing is performed on the environmental data based on a sliding average method, and a preset window size of the sliding average method is 3.

[0012] In addition, in order to achieve the above-mentioned purpose, the present application also proposes a health risk prediction device based on environmental data fusion, and the health risk prediction device based on environmental data fusion includes: a memory, a processor, and a health risk prediction program based on environmental data fusion stored on the memory and runnable on the processor. The health risk prediction program based on environmental data fusion is configured to implement the steps of the health risk prediction method based on environmental data fusion as described in any of the above items.

[0013] In addition, in order to achieve the above-mentioned purpose, the present application also proposes a storage medium, on which a health risk prediction program based on environmental data fusion is stored. When the health risk prediction program based on environmental data fusion is executed by a processor, the steps of the health risk prediction method based on environmental data fusion as described in any of the above items are implemented.

[0014] The present invention has at least the following beneficial effects: by collecting multiple different types of environmental data, the environmental data used in the present invention covers multiple fields, the types of environmental data involved are defined as target types, the collected environmental data are preprocessed according to the target types of the environmental data, and the timestamps of the environmental data of all target types are aligned by linear interpolation, so that each target type of environmental data has corresponding data at the collection time corresponding to each timestamp, and the environmental data after the aligned timestamps are further normalized to obtain normalized environmental data, so that different types of environmental data are converted to the same scale, and then risk parameters are determined based on the normalized environmental data, and risk levels are determined and output based on the risk parameters, thereby realizing early warning of health risks. Compared with environmental pollution early warning that only focuses on environmental data in a single field, the present invention combines multiple different types of environmental data, standardizes different types of environmental data through preprocessing, aligning timestamps and normalization, and integrates to determine risk parameters and risk levels, realizing the fusion of different types of environmental data, thereby comprehensively integrating multi-dimensional environmental data to make health risk early warnings that affect residents' health, thereby improving residents' physical health. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings are used to provide a further understanding of the technical solution of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the technical solution of the present invention and do not constitute a limitation to the technical solution of the present invention.

[0016] Figure 1 This is a flow chart of a first embodiment of a health risk prediction method based on environmental data fusion according to the present invention; Figure 2 This is a flow chart of a second embodiment of the health risk prediction method based on environmental data fusion according to the present invention; Figure 3This is a flow chart of a third embodiment of the health risk prediction method based on environmental data fusion according to the present invention; Figure 4 This is a flow chart of a fourth embodiment of a health risk prediction method based on environmental data fusion according to the present invention; Figure 5 This is a schematic diagram of the device structure involved in the operation of an embodiment of a health risk prediction device based on environmental data fusion according to the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0018] It should be noted that although the system diagrams illustrate functional module divisions and the flowcharts illustrate logical sequences, in certain circumstances, the steps shown or described may be performed in a sequence that differs from the module divisions in the system or the sequence in the flowcharts. The terms "first," "second," and so on, in the specification, claims, and drawings, are used to distinguish similar items and are not necessarily used to describe a specific sequence or precedence.

[0019] The embodiment of the present invention provides a health risk prediction method based on environmental data fusion. Figure 1 , a first embodiment of the health risk prediction method based on environmental data fusion is proposed. In this embodiment, the health risk prediction method based on environmental data fusion includes: Step S100 : collecting environmental data. The environmental data includes multiple different types of data. Each type of environmental data includes multiple items of data with the collection time as a timestamp. The type of environmental data is defined as a target type.

[0020] Specifically, environmental data includes data from multiple fields, such as air quality data, water quality data, and meteorological data. Air quality data refers to indoor air in residents and includes data such as PM2.5 concentration, PM10 concentration, SO2 concentration, oxygen concentration, TVOC concentration, and O3 concentration. Water quality data includes data such as TDS value. Meteorological data includes data such as temperature, humidity, wind direction, wind speed, UV index, rainfall, light intensity, and atmospheric pressure. Furthermore, environmental data may also include data related to the outdoor environment of a house, such as ambient noise, traffic and pedestrian flow, soil salinity, soil moisture, and soil nitrogen, phosphorus, and potassium content, as well as data related to furniture and electrical appliances, such as voltage, formaldehyde, radon, CO2, electromagnetic radiation, mobile phone signal strength, and Wi-Fi signal penetration through walls. For the various data types in the above examples, environmental data includes at least two of them, so that environmental data covers multiple fields. For ease of explanation, the types of environmental data involved in the method of this embodiment are defined as target types. For example, this embodiment uses the target types including TDS value, PM2.5 concentration, and voltage as an example for explanation.

[0021] By setting up sensor networks indoors and outdoors in residents' homes, the above-mentioned various types of environmental data are collected. For the collected environmental data, the collection time of each data is marked with a timestamp, and for the same type of environmental data, a corresponding data sequence is formed according to the time order of the timestamps.

[0022] Step S200 , preprocessing the environmental data according to the target type of the environmental data, aligning the timestamps of the environmental data of all target types based on linear interpolation, and normalizing the environmental data after the aligned timestamps to obtain normalized environmental data.

[0023] For environmental data of corresponding target types, preprocess the corresponding environmental data according to their target types. The preprocessing operation includes removing outliers. According to their target types, the normal data range of the corresponding environmental data can be determined, and whether each data item in the data sequence falls within the normal data range can be determined separately. If it does not fall within the range, the data item is judged to be an outlier and is removed from the data sequence. For example, for the TDS value of water quality, the normal data range is 50-1000mg / L, and 0mg / L and negative values ​​are outliers and need to be removed.

[0024] On the other hand, since different types of environmental data are collected based on different types of sensors, due to the differences between different sensors, it can be understood that the collection periods of different types of environmental data are also different. In order to perform analysis based on specific time nodes, it is necessary to align the timestamps of different types of environmental data. Specifically, the timestamp of the environmental data of one target type is determined as the benchmark timestamp. The determination method can be to select the target type with the smallest collection period, compare the timestamps of the environmental data of other target types with the benchmark timestamp, and determine whether there are missing data items. When there is no data item for the environmental data of other target types at a certain benchmark timestamp, linear interpolation is performed based on two adjacent data items close to the benchmark timestamp to obtain the data item at the benchmark timestamp, thereby achieving timestamp alignment of environmental data of all target types.

[0025] Furthermore, in order to achieve the standardization of different types of environmental data, normalization processing is required. Each target type of environmental data is normalized separately. The normalization processing is based on the maximum and minimum normalization formula, and the calculation formula is:

[0026] Where min(x) is the minimum value in the data sequence of the environmental data corresponding to the target type, and max(x) is the maximum value in the data sequence of the environmental data corresponding to the target type. Normalized environmental data is obtained through normalization processing, so that different types of environmental data can be standardized to the same scale of [0, 1] for data fusion analysis.

[0027] Step S300 , determining risk parameters based on the normalized environmental data, and determining and outputting a risk level based on the risk parameters, wherein the risk parameters represent the probability of a high risk occurring.

[0028] The risk parameter is expressed as a numerical value falling within the range of [0,1], which represents the probability of high risk. The risk level includes high risk and low risk. The probability of high risk is calculated and predicted based on the normalized environmental data, and then it is determined whether it is high risk or low risk based on the probability of high risk.

[0029] It should be noted that the risk parameters and risk levels are in correspondence with the timestamps, that is, for multiple data in the data sequence, the risk parameters and risk levels of the time nodes corresponding to each timestamp are determined respectively. When a high-risk or low-risk risk level is predicted, a health risk warning for the time node corresponding to its timestamp is output. The health risk warning includes guidance and suggestions for residents' lives. For example, when the risk level is high, residents are reminded to pay attention to indoor air circulation, and when the risk level is low, residents are reminded to keep the indoor and surrounding environments normal.

[0030] The embodiments of the present invention have at least the following beneficial effects: by collecting multiple different types of environmental data, the environmental data used in the present invention covers multiple fields, the types of environmental data involved are defined as target types, the collected environmental data are preprocessed according to the target types of the environmental data, and the timestamps of all types of environmental data are aligned by linear interpolation, so that each target type of environmental data has corresponding data at the collection time corresponding to each timestamp, and the environmental data after the aligned timestamps are further normalized to obtain normalized environmental data, so that different types of environmental data are converted to the same scale, and then risk parameters are determined based on the normalized environmental data, and risk levels are determined and output based on the risk parameters, thereby realizing early warning of health risks. Compared with environmental pollution early warning that only focuses on environmental data in a single field, the present invention combines multiple different types of environmental data, standardizes different types of environmental data through preprocessing, aligning timestamps and normalization, and integrates to determine risk parameters and risk levels, realizing the fusion of different types of environmental data, thereby comprehensively integrating multi-dimensional environmental data to make health risk early warnings that affect residents' health, thereby improving residents' physical health.

[0031] Furthermore, in this embodiment, before the step of normalizing the environmental data after aligning the timestamps, the following steps are further included: For the environmental data of each target type, noise reduction is performed on the environmental data based on the sliding average method, and the preset window size of the sliding average method is 3.

[0032] In this embodiment, in addition to the aforementioned preprocessing, interpolation and timestamp alignment, and normalization, environmental data processing also includes noise reduction. This noise reduction process is performed between the interpolation and timestamp alignment operations and normalization to eliminate the effects of sensor noise or accidental fluctuations, resulting in smoother, more reliable data. For each target type of environmental data sequence, a sliding average is sequentially performed using a preset window size of 3. After the sliding average, the environmental data is subjected to normalization.

[0033] Furthermore, based on the above embodiment, a second embodiment of the health risk prediction method based on environmental data fusion is proposed. Figure 2 In this embodiment, the step of determining the risk parameter based on the normalized environmental data includes: Step S311, obtaining the weight parameter of the target type.

[0034] Step S312: For the normalized environment data with the same timestamp, a weighted average value is calculated according to the weight parameter and the normalized environment data, and the weighted average value is defined as a sub-parameter of the risk parameter.

[0035] In this embodiment, different weights are assigned to each target type of environmental data and weighted averaged to achieve the fusion of multiple different types of environmental data. Specifically, the weight parameter is related to the target type. According to the target type involved, the preset weight parameter corresponding to each target type is obtained. For example, the target types include TDS value, PM2.5 concentration and voltage, and the weight parameters are 0.4, 0.4 and 0.2 respectively. For the environmental data corresponding to a certain timestamp, the weighted average value can be calculated by the following formula:

[0036] in, 、 and The weight parameters corresponding to TDS value, PM2.5 concentration and voltage respectively, is the normalized TDS value corresponding to the timestamp, is the normalized PM2.5 concentration corresponding to the timestamp, is the normalized voltage corresponding to the timestamp.

[0037] In other embodiments, the weight parameters are also related to the combination of target types. If environmental data of target types with different combinations are analyzed, the weights of environmental data of the same type may also change. For example, different from the above example, the target types include TDS value, PM2.5 concentration and electromagnetic radiation. Therefore, when obtaining the preset weight parameters, it is also necessary to obtain matching weight parameters based on the combination of target types and the target type of the environmental data.

[0038] The calculated weighted average falls within the range of [0,1]. The larger the value, the greater the possibility of abnormality. Therefore, it can represent the probability of high risk and be defined as a sub-parameter of the risk parameter.

[0039] After the weighted average value is calculated, step S320 may be executed to determine and output the risk level according to the risk parameters.

[0040] By using weighted averaging to fuse different types of environmental data to analyze the probability of high risk, it is beneficial to analyze the differentiation of different types of environmental data according to weights, improve the accuracy of risk level prediction, and thus improve the physical health of residents.

[0041] Furthermore, based on the above embodiments, a third embodiment of the health risk prediction method based on environmental data fusion is proposed. Figure 3 In this embodiment, the step of determining the risk parameter based on the normalized environmental data includes: Step S313: Acquire event status information and conditional probability information that match the target type, and form a Bayesian network of the target type based on the event status information and conditional probability information.

[0042] In step S314 , based on the Bayesian network, the Bayesian probability of high risk is determined according to the normalized environmental data, and the Bayesian probability is defined as a sub-parameter of the risk parameter.

[0043] A Bayesian network is a graphical model based on probabilistic reasoning, used to represent the dependency relationship between variables. It consists of a directed acyclic graph and a conditional probability table. The directed acyclic graph is used to represent the dependency relationship between variables, and is represented by several nodes connected by several directed edges. The node represents a variable, which is also equivalent to an event. Two nodes connected by a directed edge represent that there is an association between the corresponding two events, and the conditional probability table is used to describe the probabilistic relationship between a node and its parent node, including the probability value of the current node taking each possible value given the state of the parent node.

[0044] In this embodiment, the event state information includes the various nodes required by the Bayesian network, which specifically includes the normal and abnormal states of the environmental data of each target type, and the final risk level assessment of the two states of high risk and low risk. It also includes the correlation relationship between the state of the environmental data and the impact of the risk level, and the conditional probability information includes the specific probability of the state of the environmental data affecting the risk level. Both the event state information and the conditional probability information are related to the combination of target types. Therefore, by obtaining the event state information and conditional probability information that match the target type, a Bayesian network of the target type can be formed based on it. The state of the parent node is determined based on the numerical value of the normalized environmental data, and the Bayesian probability can be determined by referring to the corresponding conditional probability information. The Bayesian probability represents the probability of high risk and is defined as a sub-parameter of the risk parameter.

[0045] For example, the target types include TDS value, PM2.5 concentration, and voltage. TDS value includes normal and abnormal states, PM2.5 concentration includes low and high states, and voltage includes normal and abnormal states. In addition, it is set that when the TDS value is abnormal and the PM2.5 concentration is high, the high risk probability increases, and when the voltage is abnormal, the high risk probability increases. On the other hand, conditional probabilities are also set:

[0046] Among them, R indicates the risk level. By inputting the numerical value of the normalized environmental data and determining the status of the three environmental data, the corresponding Bayesian probability can be inferred based on the set conditional probability.

[0047] After the Bayesian probability is calculated, step S320 may be executed to determine and output the risk level according to the risk parameters.

[0048] By using Bayesian networks to fuse different types of environmental data to analyze the probability of high risk, it is beneficial to use probabilistic reasoning to analyze the dependencies between the states of environmental data and between environmental data and risk levels, thereby improving the accuracy of risk level predictions and improving the physical health of residents.

[0049] Furthermore, based on the above embodiments, a fourth embodiment of the health risk prediction method based on environmental data fusion is proposed. Figure 4 In this embodiment, the step of determining the risk parameter based on the normalized environmental data includes: Step S315: Obtain a logistic regression prediction model for the target type.

[0050] Step S316: input the normalized environmental data into the logistic regression prediction model, determine the model prediction probability according to the output result of the logistic regression prediction model, and define the model prediction probability as a sub-parameter of the risk parameter.

[0051] In this embodiment, the logistic regression prediction model is used to solve the binary classification problem of risk level to predict whether the risk level is high risk or low risk. The obtained logistic regression prediction model also needs to match the combination of target types. For different combinations of target types, the corresponding logistic regression prediction model should be trained separately for prediction. The normalized environmental data after normalization in the above embodiment is input into the logistic regression prediction model, and the probability of it being high risk is calculated according to its logistic regression formula. The specific calculation formula is:

[0052] in, The output is 1, that is, the probability of the output being high risk. Defined as the model prediction probability, is the bias term of the model, i.e. the intercept, is the regression coefficient of each indicator, are normalized environmental data with the same timestamp.

[0053] Based on the above calculation formula, the output result of the model prediction probability can be obtained by inputting normalized environmental data, and the model prediction probability is also defined as a sub-parameter of the risk parameter.

[0054] After the model prediction probability is calculated, step S320 may be executed to determine and output the risk level according to the risk parameters.

[0055] By using the logistic regression model to integrate different types of environmental data to analyze the probability of high risk, it is beneficial to analyze the binary problem of whether the risk level is high risk or low risk, improve computing efficiency, and also help improve the accuracy of risk level prediction, thereby improving the physical health of residents.

[0056] Furthermore, in this embodiment, the health risk prediction method based on environmental data fusion further includes: Get the training environment data that matches the target type.

[0057] The preset model is trained according to the training environment data, and the parameters of the preset model are adjusted based on minimizing the logarithmic loss function until the prediction accuracy of the preset model for the training environment data is greater than or equal to the preset accuracy, thereby obtaining a logistic regression prediction model, wherein the parameters of the preset model include the intercept and the regression coefficient.

[0058] Before using the logistic regression prediction model of the target type, it is necessary to first train the model. Specifically, obtain training environment data that matches the target type. The training environment data includes multiple data items of the same target type, and each timestamp data item is also provided with a label to mark its risk level. The training environment data is divided into a training set and a test set. The preset model is trained based on the training set. The parameters of the preset model include the above-mentioned intercept and regression coefficient. Before starting training, the intercept and regression coefficient are initialized to the preset values. When the preset model is trained, its parameters are adjusted based on minimizing the logarithmic loss function until the prediction accuracy of the preset model for the test set is greater than or equal to the preset accuracy. The values ​​of the intercept and regression coefficient are determined, and the training of the logistic regression prediction model of the target type is completed.

[0059] By training and optimizing the logistic regression prediction model based on minimizing the logarithmic loss function, the prediction accuracy of the logistic regression prediction model is improved, the accuracy of risk level prediction is improved, and thus the physical health of residents is improved.

[0060] Furthermore, based on the above embodiments, a fifth embodiment of the health risk prediction method based on environmental data fusion is proposed. In this embodiment, the step of determining and outputting the risk level based on the risk parameters includes: When the risk parameter includes a sub-parameter, the risk level is determined based on the sub-parameter, and a health risk warning corresponding to the risk level is output.

[0061] When a risk parameter includes more than one sub-parameter, the risk level is determined by voting based on all sub-parameters, and a health risk warning corresponding to the risk level is output.

[0062] Based on the above embodiment, the method of determining risk parameters based on normalized environmental data includes calculating a weighted average, using a Bayesian network, and using a logistic regression prediction model, which respectively obtain sub-parameters of three risk parameters: weighted average, Bayesian probability, and model prediction probability. Among them, the weighted average can analyze the differentiation of different types of environmental data according to the weights, the Bayesian network can use probabilistic reasoning to analyze according to the dependency between the states of the environmental data and between the environmental data and the risk level, and the logistic regression prediction model is suitable for analyzing the binary differentiation problem of whether the risk level is high risk or low risk. It obtains risk parameters from different angles and with different advantages.

[0063] In this embodiment, the method for determining the risk parameter based on the normalized environmental data can be any one of weighted average, Bayesian network, and logistic regression prediction model, or all of them. The selected methods must be an odd number so that the final result can be determined. In other embodiments, if there are other methods that can determine the risk parameter based on the normalized environmental data, they can also be included in the use to ensure that the final selected methods total an odd number. When only one method is selected, the risk parameter includes a sub-parameter, and the risk level is directly determined based on the sub-parameter, and then the health risk warning corresponding to the risk level is output; when more than one method is selected, in this embodiment, it is equivalent to the sub-parameter including the weighted average, Bayesian probability, and model prediction probability at the same time, and the risk level corresponding to each sub-parameter is determined separately, and then it is voted on whether it is high risk or low risk, and the result with the higher number of votes is determined as the final risk level, and then the health risk warning corresponding to the risk level is output.

[0064] By integrating multiple sub-parameters of risk parameters and using voting to determine the final risk level, the present invention can integrate different analysis methods based on the integration of different types of environmental data, and combine the advantages of different analysis methods to obtain the optimal health risk prediction results, thereby improving the physical health of residents.

[0065] Furthermore, in this embodiment, the step of determining the risk level according to the sub-parameters includes: When the sub-parameter is greater than or equal to the preset probability, the risk level is determined to be high risk.

[0066] When the sub-parameter is less than the preset probability, the risk level is determined to be low risk.

[0067] Since the risk parameter is expressed as a value falling within the range of [0,1], representing the probability of high risk, the various sub-parameters are similar. The weighted average, Bayesian probability and model prediction probability are all values ​​falling within the range of [0,1]. Therefore, a preset probability is set, for example 0.5. When the sub-parameter is greater than or equal to the preset probability, the risk level is judged to be high risk. When the sub-parameter is less than the preset probability, the risk level is judged to be low risk.

[0068] In other embodiments, due to differences in calculation methods, different sub-parameters may not be suitable for using the same indicators to judge their risk levels. Therefore, corresponding preset probabilities can be set as judgment indicators for different sub-parameters. That is, before comparing the sub-parameters with the preset probabilities, a step is added to determine the corresponding preset probabilities based on the sub-parameters, so that each sub-parameter can be accurately judged, thereby improving the accuracy of the risk level.

[0069] On the other hand, reference Figure 5 , Figure 5 This is a schematic diagram of the device structure of a health risk prediction device based on environmental data fusion.

[0070] A device for predicting health risks based on environmental data fusion is provided, comprising: a processor and a memory; wherein the memory is configured to store a computer-readable program. When the computer-readable program is executed by the processor, the processor implements a method for predicting health risks based on environmental data fusion as described in any of the above technical solutions.

[0071] Those skilled in the art will appreciate that all or some of the steps and systems disclosed above can be implemented as software, firmware, hardware, or any combination thereof. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on computer-readable media, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is well known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVDs) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. As is well known to those skilled in the art, communication media typically embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0072] In addition, an embodiment of the present invention also proposes a storage medium, which stores a health risk prediction program based on environmental data fusion. When the health risk prediction program based on environmental data fusion is executed by a processor, the relevant steps of any embodiment of the health risk prediction method based on environmental data fusion as described above are implemented.

[0073] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can, for example, be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0074] It should be understood that in this application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0075] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0076] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0077] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0078] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0079] Although the description of the present application has been quite detailed and specifically describes several embodiments, it is not intended to be limited to any of these details or embodiments or any particular embodiment, but should be considered to provide a broad possible interpretation of these claims by reference to the appended claims, taking into account the prior art, so as to effectively cover the intended scope of the present application. In addition, the above description of the present application is based on the embodiments foreseeable by the inventors, which is intended to provide a useful description, and those non-substantial changes to the present application that have not yet been foreseen may still represent equivalent changes to the present application.

Claims

1. A health risk prediction method based on environmental data fusion, characterized in that: The health risk prediction method based on environmental data fusion includes: Collecting environmental data, the environmental data includes multiple different types of data, each type of environmental data includes multiple data with the collection time as a timestamp, and defining the type of the environmental data as a target type; Preprocessing the environmental data according to the target type of the environmental data, aligning the timestamps of the environmental data of all the target types based on linear interpolation, and normalizing the environmental data after the timestamps are aligned to obtain normalized environmental data; A risk parameter is determined according to the normalized environmental data, and a risk level is determined and output according to the risk parameter, wherein the risk parameter represents a probability of occurrence of a high risk.

2. The health risk prediction method based on environmental data fusion according to claim 1 is characterized in that: The step of determining the risk parameter according to the normalized environmental data comprises: Obtaining a weight parameter of the target type; For the normalized environment data with the same timestamp, a weighted average value is calculated according to the weight parameter and the normalized environment data, and the weighted average value is defined as a sub-parameter of the risk parameter.

3. The health risk prediction method based on environmental data fusion according to claim 1 is characterized in that: The step of determining the risk parameter according to the normalized environmental data comprises: Acquire event state information and conditional probability information matching the target type, and form a Bayesian network of the target type according to the event state information and the conditional probability information; Based on the Bayesian network, a Bayesian probability of high risk is determined according to the normalized environmental data, and the Bayesian probability is defined as a sub-parameter of the risk parameter.

4. The health risk prediction method based on environmental data fusion according to claim 1 is characterized in that: The step of determining the risk parameter according to the normalized environmental data comprises: Obtaining a logistic regression prediction model for the target type; The normalized environmental data is input into the logistic regression prediction model, the model prediction probability is determined according to the output result of the logistic regression prediction model, and the model prediction probability is defined as a sub-parameter of the risk parameter.

5. The health risk prediction method based on environmental data fusion according to claim 4 is characterized in that: The health risk prediction method based on environmental data fusion also includes: Acquire training environment data matching the target type; A preset model is trained according to the training environment data, and the parameters of the preset model are adjusted based on minimizing the logarithmic loss function until the prediction accuracy of the preset model for the training environment data is greater than or equal to the preset accuracy, thereby obtaining the logistic regression prediction model, wherein the parameters of the preset model include an intercept and a regression coefficient.

6. The health risk prediction method based on environmental data fusion according to claim 1 is characterized in that: The step of determining and outputting the risk level according to the risk parameter comprises: When the risk parameter includes a sub-parameter, determining the risk level according to the sub-parameter, and outputting a health risk warning corresponding to the risk level; When the risk parameter includes more than one sub-parameter, the risk level is determined by voting based on all the sub-parameters, and a health risk warning corresponding to the risk level is output.

7. The health risk prediction method based on environmental data fusion according to claim 6 is characterized in that: The step of determining the risk level according to the sub-parameters comprises: When the sub-parameter is greater than or equal to a preset probability, the risk level is determined to be high risk; When the sub-parameter is less than the preset probability, the risk level is determined to be low risk.

8. The health risk prediction method based on environmental data fusion according to claim 1 is characterized in that: Before the step of normalizing the environmental data after aligning the timestamp, the method further includes: For the environmental data of each target type, noise reduction processing is performed on the environmental data based on a sliding average method, and a preset window size of the sliding average method is 3.

9. A health risk prediction device based on environmental data fusion, characterized in that: The health risk prediction device based on environmental data fusion includes: a memory, a processor, and a health risk prediction program based on environmental data fusion stored in the memory and executable on the processor, wherein the health risk prediction program based on environmental data fusion is configured to implement the steps of the health risk prediction method based on environmental data fusion as described in any one of claims 1 to 8.

10. A storage medium, characterized in that: The storage medium stores a health risk prediction program based on environmental data fusion, and when the health risk prediction program based on environmental data fusion is executed by a processor, the steps of the health risk prediction method based on environmental data fusion as described in any one of claims 1 to 8 are implemented.

Images