Probability forecast method, device, storage medium and electronic device

By acquiring the observation data set and the forecast data set, calculating the conditional probability and occurrence probability of each forecast scheme, and targeting the occurrence probability under the forecast result set of the target event, solving the problem of low probability forecast accuracy in the prior art. Achieve higher forecast accuracy.

CN118690144BActive Publication Date: 2024-11-193CLEAR SCI & TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411174166.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-26
Publication Date
2024-11-19
Estimated Expiration
2044-08-26

AI Technical Summary

Technical Problem

The existing probability forecasting technology has low accuracy in air quality forecasting, and it has failed to effectively consider the performance differences of different forecasting plans, resulting in insufficient accuracy of forecast results.

Method used

By obtaining the forecast data set of observation data sets and each forecast plan, calculate the conditional probability and occurrence probability of each forecast plan within different time ranges, and targetedly determine the probability of occurrence of target events under the forecast result set, improving the accuracy of forecasts .

Benefits of technology

It improves the accuracy of probability forecasting, can predict the probability of target events more accurately, and enhances the reliability of forecast results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118690144B_ABST
    Figure CN118690144B_ABST
Patent Text Reader

Abstract

The present invention provides a probability forecasting method, device, storage medium and electronic device, wherein the method comprises: determining the forecast condition probability set of each forecast scheme in each second time range in at least one second time range based on the observation data set and the forecast data set corresponding to each forecast scheme; calculating the observation occurrence probability set in each first time range based on the observation data set; and calculating the forecast probability set of each forecast scheme in each first time range based on the forecast data set corresponding to each forecast scheme; determining the occurrence probability of each target event in at least one target event under the event forecast result set based on the observation occurrence probability set in each first time range, the forecast probability set of each forecast scheme in each first time range and the forecast condition probability set of each forecast scheme in each second time range. The embodiment of the present invention can effectively improve the accuracy of probability forecasting.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a probability forecasting method, device, storage medium and electronic equipment. Background Art

[0002] At present, probabilistic forecasting has been widely used in various scenarios, such as air quality forecasting, precipitation level forecasting, wind force level forecasting, and primary pollutant forecasting. Among them, when probabilistic forecasting is applied to air quality forecasting, the so-called probabilistic forecasting is a forecast of the probability of a certain meteorological element or weather element appearing within a certain range of values ​​in the future. That is to say, probabilistic forecasting not only predicts the air quality level of a certain day, but also gives the expected probability of occurrence. However, related technologies usually use multiple numerical methods (also called multiple set members) to predict events separately, and then perform statistics based on the results of different methods to obtain the probability of each event. For example, if there are 4 methods in total, 3 of which predict that the air quality level of a certain place tomorrow is slightly polluted, then the forecast result is that the air quality level of a certain place tomorrow is slightly polluted, and the probability of its occurrence is 75%; in this regard, each method has the same impact on probability statistics, that is, the weight of each method is the same, and the different performances of different methods are not considered, resulting in low accuracy of probabilistic forecasting. Based on this, there is currently no good solution to improve the accuracy of probabilistic forecasting. Summary of the invention

[0003] In view of this, the embodiments of the present invention provide a probability forecasting method, device, storage medium and electronic device to solve the problems of low accuracy of probability forecasting in related technologies; that is, the embodiments of the present invention can use the observation data set and the forecast data set corresponding to each forecast scheme to specifically consider the forecast probability and forecast conditional probability under different forecast schemes, that is, specifically consider the forecast probability and forecast conditional probability of each forecast event when forecasting events under different forecast schemes, so as to accurately perform probability forecasting and obtain probability forecasting results with higher accuracy (that is, the probability of occurrence of each target event under the event forecast result set), which can effectively improve the accuracy of probability forecasting.

[0004] According to one aspect of an embodiment of the present invention, a probability forecasting method is provided, the method comprising:

[0005] Obtain an observation data set and a forecast data set corresponding to each forecast scheme in N forecast schemes, wherein the observation data set includes: an observation event data set of a target area in each first time range in at least one first time range, and a forecast data set includes: a forecast event data set of the target area in each first time range under a corresponding forecast scheme; wherein one forecast event data is obtained by event forecasting through the forecast scheme corresponding to the forecast data set, and N is a positive integer;

[0006] Based on the observation data set and the forecast data set corresponding to each forecast scheme, determine the forecast condition probability set of each forecast scheme in each second time range in at least one second time range, wherein the forecast condition probability set of a forecast scheme in a second time range is: the forecast condition probability set obtained by the corresponding forecast scheme for forecasting events in the target area in the corresponding second time range;

[0007] Based on the observation data set, calculating the observation occurrence probability set within each first time range; and based on the forecast data set corresponding to each forecast scheme, calculating the forecast probability set of each forecast scheme within each first time range;

[0008] Based on the observed occurrence probability sets within the respective first time ranges, the forecast probability sets of the respective forecast schemes within the respective first time ranges, and the forecast conditional probability sets of the respective forecast schemes within the respective second time ranges, the occurrence probability of each target event in at least one target event under the event forecast result set is determined, and the event forecast result set includes the event forecast results obtained by each forecast scheme performing event forecasts on the target area within the third time range.

[0009] According to another aspect of an embodiment of the present invention, a probability forecasting device is provided, the device comprising:

[0010] An acquisition unit is used to acquire an observation data set and a forecast data set corresponding to each forecast scheme in N forecast schemes, wherein the observation data set includes: an observation event data set of a target area in each first time range in at least one first time range, and a forecast data set includes: a forecast event data set of the target area in each first time range under a corresponding forecast scheme; wherein one forecast event data is obtained by forecasting an event through the forecast scheme corresponding to the forecast data set, and N is a positive integer;

[0011] A processing unit, configured to determine, based on the observation data set and the forecast data set corresponding to each forecast scheme, a forecast condition probability set of each forecast scheme in each second time range in at least one second time range, wherein the forecast condition probability set of a forecast scheme in a second time range is: a forecast condition probability set obtained by the corresponding forecast scheme performing event forecasting on the target area in the corresponding second time range;

[0012] The processing unit is further configured to calculate the observation occurrence probability set within each first time range based on the observation data set; and calculate the forecast probability set of each forecast scheme within each first time range based on the forecast data set corresponding to each forecast scheme;

[0013] The processing unit is also used to determine the occurrence probability of each target event in at least one target event under an event forecast result set based on the observed occurrence probability set within the each first time range, the forecast probability set of each forecast scheme within the each first time range, and the forecast conditional probability set of each forecast scheme within the each second time range. The event forecast result set includes the event forecast results obtained by each forecast scheme for event forecasting the target area within the third time range.

[0014] According to another aspect of an embodiment of the present invention, an electronic device is provided, comprising a processor and a memory storing a program, wherein the program comprises instructions, which, when executed by the processor, enable the processor to perform the above-mentioned method.

[0015] According to another aspect of an embodiment of the present invention, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable a computer to execute the above-mentioned method.

[0016] The embodiment of the present invention can determine the forecast condition probability set of each forecast scheme in each second time range in at least one second time range based on the observation data set and the forecast data set corresponding to each forecast scheme after obtaining the observation data set and the forecast data set corresponding to each forecast scheme, respectively. The forecast condition probability set of a forecast scheme in a second time range is: the forecast condition probability set obtained by the corresponding forecast scheme for the target area in the corresponding second time range through event forecasting; the observation data set includes: the observation event data set of the target area in each first time range in at least one first time range, and a forecast data set includes: the forecast event data set of the target area in each first time range under the corresponding forecast scheme; wherein, a forecast event data is obtained by forecasting an event through the forecast scheme corresponding to the forecast data set, and N is a positive integer. Accordingly, the observation occurrence probability set in each first time range can be calculated based on the observation data set; and the forecast probability set of each forecast scheme in each first time range can be calculated based on the forecast data set corresponding to each forecast scheme. Based on this, the occurrence probability of each target event in at least one target event under the event forecast result set can be determined based on the observed occurrence probability set within each first time range, the forecast probability set of each forecast scheme within each first time range, and the forecast conditional probability set of each forecast scheme within each second time range. The event forecast result set includes the event forecast results obtained by each forecast scheme for forecasting the target area within the third time range. It can be seen that the embodiment of the present invention can specifically consider the forecast probability and forecast conditional probability under different forecast schemes through the observation data set and the forecast data set corresponding to each forecast scheme, that is, specifically consider the forecast probability and forecast conditional probability of each forecast event when different forecast schemes are used to forecast events, so as to accurately perform probability forecasting and obtain a probability forecasting result with higher accuracy (that is, the occurrence probability of each target event under the event forecast result set), which can effectively improve the accuracy of probability forecasting. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Further details, features and advantages of the invention are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which:

[0018] Figure 1 A schematic flow chart of a probabilistic forecasting method according to an exemplary embodiment of the present invention is shown;

[0019] Figure 2 A schematic flow chart of another probabilistic forecasting method according to an exemplary embodiment of the present invention is shown;

[0020] Figure 3 A schematic block diagram of a probability forecasting device according to an exemplary embodiment of the present invention is shown;

[0021] Figure 4 A block diagram of an exemplary electronic device that can be used to implement an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0022] Embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are only for exemplary purposes and are not intended to limit the scope of protection of the present invention.

[0023] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0024] The term "including" and its variations used in this document are open inclusions, that is, "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments". Relevant definitions of other terms will be given in the following description. It should be noted that the concepts of "first", "second", etc. mentioned in the present invention are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0025] It should be noted that the modifications of "one" and "plurality" mentioned in the present invention are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0026] The names of the messages or information exchanged between multiple devices in the embodiments of the present invention are only used for illustrative purposes, and are not used to limit the scope of these messages or information.

[0027] It should be noted that the execution subject of the probability forecasting method provided in the embodiment of the present invention may be one or more electronic devices, and the present invention does not limit this; wherein, the electronic device may be a terminal (i.e., a client) or a server, then when the execution subject includes multiple electronic devices, and the multiple electronic devices include at least one terminal and at least one server, the probability forecasting method provided in the embodiment of the present invention may be jointly executed by the terminal and the server. Accordingly, the terminal mentioned here may include but is not limited to: smart phones, tablet computers, laptops, desktop computers, smart watches, smart voice interaction devices, etc.; the server mentioned here may be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud services, cloud databases, cloud computing (cloud computing), cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and basic cloud computing services such as big data and artificial intelligence platforms, etc.

[0028] Based on the above description, an embodiment of the present invention proposes a probability forecasting method, which can be executed by the electronic device (terminal or server) mentioned above; or, the probability forecasting method can be executed by the terminal and the server together. For the sake of convenience, the following description will be given by taking the electronic device executing the probability forecasting method as an example; Figure 1 As shown, the probability forecasting method may include the following steps S101-S104:

[0029] S101, obtaining an observation data set and a forecast data set corresponding to each forecast scheme in N forecast schemes, wherein the observation data set includes: an observation event data set of a target area within each first time range in at least one first time range, and a forecast data set includes: a forecast event data set of the target area within each first time range under a corresponding forecast scheme; wherein one forecast event data is obtained by event forecasting through the forecast scheme corresponding to the forecast data set, and N is a positive integer.

[0030] In an embodiment of the present invention, the N forecasting schemes may include but are not limited to at least one of the following: forecasting schemes for each time effect in at least one time effect and each grid division method in at least one grid division method (in this case, one forecasting scheme may be a forecasting scheme for one time effect and one grid division method for each ensemble forecasting model in at least one ensemble forecasting model for each time effect and each grid division method (in this case, one forecasting scheme may be a forecasting scheme for one ensemble forecasting model for one time effect and one grid division method), etc.; the embodiment of the present invention does not limit this. Optionally, an ensemble forecasting model may include multiple ensemble members, and one ensemble member may be a numerical model or a statistical model, etc., which is not limited in the embodiment of the present invention. Optionally, a time limit can be used to indicate the time interval between the forecast time of an event and the corresponding occurrence time of the corresponding event (such as one day, two days or three days, etc.); illustratively, a time limit can be a time limit with a time interval of 1 day (in this case, an event forecast can be made for the target area on the second day), or a time limit can be a time limit with a time interval of 2 days (in this case, an event forecast can be made for the target area on the third day), and so on. It should be understood that at least one time limit and at least one grid division format can be set according to experience or actual needs, and the embodiment of the present invention is not limited to this. Optionally, N forecasting schemes can be set according to experience or actual needs, and the embodiment of the present invention is not limited to this.

[0031] Optionally, an event (i.e., a forecast event or an observation event, etc.) may refer to an air quality level event (such as light pollution, heavy pollution, etc.) or a primary pollutant event (such as PM10 (inhalable particulate matter), PM2.5 (fine particulate matter), NO2 (nitrogen dioxide), O3 (ozone), SO2 (sulfur dioxide), CO (nitric oxide), etc.), etc. In this case, at least one numerical model may include but is not limited to at least one of the following: CMAQ (Community Multiscale Air Quality, an air quality forecasting and assessment system), NAQPMS (Nested Air Quality Prediction Modeling System, a nested grid air quality forecasting model system), CAMx (an atmospheric pollutant calculation model based on atmospheric chemistry for processes such as ozone and particulate matter), etc.; the embodiment of the present invention is not limited to this. Optionally, an event may also refer to a precipitation level event (such as light rain, moderate rain or sunny, etc.) or a wind level event (such as no wind, light wind, breeze, strong wind, etc.), etc. In this case, at least one numerical mode may include but is not limited to at least one of the following: Ecmwf (a weather forecast system), Grapes (another weather forecast system), etc.; the embodiment of the present invention is not limited to this.

[0032] Among them, a first time range may correspond to a first time subordinate indicator, and a first time range may include: at least one time sub-range of the corresponding first time subordinate indicator within the first historical time range. Optionally, a first time subordinate indicator may be a month (such as January or February, etc.), in which case the subordinate indicator duration of the first time subordinate indicator is a month; or, a first time subordinate indicator may also be a quarter (such as the first quarter or the second quarter, etc.), in which case the subordinate indicator duration of the first time subordinate indicator is a quarter, and so on; the embodiment of the present invention does not limit this. Correspondingly, the first historical time range may refer to W years before the current system time; for example, the first historical time range may include the three-year time range of 2020, 2022 and 2023, and may also include the five-year time range of 2019-2023, and so on. It should be understood that the embodiment of the present invention does not limit the value of W and the specific time range of the first historical time range. Optionally, at least one first time range may include each first time range in the first historical time range (such as the first time range corresponding to each month in 12 months, or the first time range corresponding to each quarter in 4 quarters), and may also include the first time range in the first historical time range that matches the third time range, and P first time ranges corresponding to the second time range that matches the third time range (optionally, at this time, the first time range that matches the third time range and the P first time ranges corresponding to the second time range that matches the third time range can be first determined from the first historical time range), P is a positive integer, and so on; the embodiment of the present invention is not limited to this.

[0033] For example, assuming that the first time subordinate indicator corresponding to a first time range is January, and the first historical time range is 2021-2023, then the first time range corresponding to the first time subordinate indicator of January may include three time sub-ranges: January 2021, January 2022, and January 2023. For another example, assuming that the first time subordinate indicator corresponding to a first time range is the first quarter, and the first historical time range is 2021-2023, then the first time range corresponding to the first time subordinate indicator of the first quarter may include three time sub-ranges: the first quarter of 2021, the first quarter of 2022, and the first quarter of 2023, and so on.

[0034] In the embodiment of the present invention, the observed event data is obtained based on the observed event (which can be an observed event, or can be referred to as an actual event), and an observed event can be any event in the event set, that is, multiple observed events can include each event in the event set, and an observed event can be any observed event in multiple observed events. Optionally, when an event refers to an air quality level event, the event set may include but is not limited to: excellent air quality, good air quality, light pollution, moderate pollution, heavy pollution, and severe pollution; when an event refers to a primary pollutant event, the event set may include but is not limited to: inhalable particulate matter, fine particulate matter, nitrogen dioxide, ozone, sulfur dioxide, and nitrogen monoxide; when an event refers to a precipitation level event, the event set may include but is not limited to: light rain, moderate rain, heavy rain, rainstorm, extremely heavy rainstorm, and sunny; when an event refers to a wind force level event, the event set may include but is not limited to: no wind, light wind, breeze, gentle wind, strong wind, strong wind, and gale; the embodiment of the present invention does not limit this. Optionally, an observation event data may be an observed event (i.e., an observation event), or the number of occurrences of a certain observation event within the corresponding first time range, or the number of occurrences of a certain observation event within a time sub-range within the corresponding first time range, etc., which is not limited in the embodiments of the present invention. Optionally, a forecast event data may be an event forecasted by a forecast scheme (i.e., a forecast event, a forecast event may be any event in an event set, i.e., multiple forecast events may include each event in the event set, and a forecast event may be any forecast event among multiple forecast events), or the number of forecasts of a certain forecast event by a forecast scheme within the corresponding first time range, or the number of forecasts of a certain forecast event by a forecast scheme within a time sub-range within the corresponding first time range, etc., which is not limited in the embodiments of the present invention.

[0035] Exemplarily, taking the observed event data set as an example, assuming that the first time range 1 includes time sub-range 1 (such as January 2022) and time sub-range 2 (such as January 2023), the event set includes event 1 and event 2, and the events observed in time sub-range 1 include 13 events 1 and 18 events 2, and the events observed in time sub-range 2 include 11 events 1 and 20 events 2, then when an observed event data is an observed event, the observed event data set in the first time range 1 may include the events observed in time sub-range 1, and the events observed in time sub-range 2. ; When an observation event data is the number of occurrences of a certain observation event within the corresponding first time range, the observation event data set within the first time range 1 may include: the number of occurrences of event 1 within the first time range 1 (i.e. 24), and the number of occurrences of event 2 within the first time range 1 (i.e. 38); when an observation event data is the number of occurrences of a certain observation event within a time sub-range within the corresponding first time range, the observation event data set within the first time range 1 may include: the number of occurrences of event 1 and the number of occurrences of event 2 within time sub-range 1, the number of occurrences of event 1 and the number of occurrences of event 2 within time sub-range 2, and so on.

[0036] Optionally, the above observation data set and the forecast data set corresponding to each forecast scheme may be obtained in the following ways, but not limited to:

[0037] The first acquisition method: the electronic device can obtain the observation download link of the observation data set and the forecast download link of the forecast data set corresponding to each forecast scheme, and use the data set downloaded based on the observation download link as the observation data set, and use the data set downloaded based on the forecast download link of the forecast data set corresponding to each forecast scheme as the forecast data set corresponding to the corresponding forecast scheme. Alternatively, the electronic device can obtain the integrated data download link, and the data set corresponding to the integrated data download link includes the observation data set and the forecast data set corresponding to each forecast scheme, then the electronic device can download according to the integrated data download link to obtain the observation data set and the forecast data set corresponding to each forecast scheme.

[0038] The second acquisition method: the electronic device stores observation data sets corresponding to each of the multiple regions and forecast data sets corresponding to each forecast scheme, then the electronic device can determine the target region from the multiple regions, and obtain the observation data set corresponding to the target region and the forecast data set corresponding to each forecast scheme, and so on.

[0039] It should be noted that the above-mentioned target area can be a national area, an area where a province is located, an area where a city is located, etc.; the embodiment of the present invention is not limited to this.

[0040] S102, based on the observation data set and the forecast data set corresponding to each forecast scheme, determine the forecast condition probability set of each forecast scheme in each second time range in at least one second time range, and the forecast condition probability set of a forecast scheme in a second time range is: the forecast condition probability set obtained by the corresponding forecast scheme for event forecasting for the target area in the corresponding second time range.

[0041] Among them, a second time range may include: at least one time sub-range of the corresponding second time subordinate indicator within the second historical time range, and the second time subordinate indicator may be a month or a quarter, etc., which is not limited by the embodiment of the present invention; optionally, the subordinate indicator duration (such as a month or a quarter, etc.) of the second time subordinate indicator may be the same as or different from the subordinate indicator duration of the first time subordinate indicator, which is not limited by the embodiment of the present invention. Correspondingly, the second historical time range may refer to V years before the current system time, and the embodiment of the present invention does not limit the value of V and the specific time range of the second historical time range. Optionally, the second historical time range may be equal to the first historical time range, or may be smaller than the first historical time range (that is, the second historical time range is a partial time range in the first historical time range), etc., which is not limited by the embodiment of the present invention. Optionally, at least one second time range may include each second time range in the second historical time range (such as the second time range corresponding to each month in 12 months, etc.), or may only include the second time range that matches the third time range (optionally, at this time the second time range that matches the third historical time range can be first determined from the second historical time range), etc. The embodiments of the present invention are not limited to this.

[0042] Accordingly, each second time range in at least one second time range may correspond to each first time range in at least one first time range, that is, one second time range may be the same as one first time range; or, one second time range may include P first time ranges in at least one first time range, P is a positive integer, and P is less than or equal to the number of first time ranges in at least one first time range, etc. Based on this, one second time range may correspond to P first time ranges.

[0043] Optionally, an embodiment of the present invention may preferably perform calculations when the second historical time range is the same as the first historical time range, thereby fully considering the observation data set and the forecast data set corresponding to each forecast scheme when calculating the forecast conditional probability set, so as to improve the accuracy of the forecast conditional probability set.

[0044] For example, assuming that the first historical time range and the second historical time range are both 2021-2023, a first time subordinate indicator corresponding to a first time range is January, and a corresponding second time subordinate indicator corresponding to a second time range is also January, the first time range corresponding to the first time subordinate indicator being January and the second time range corresponding to the second time subordinate indicator being January can both include: January 2021, January 2022 and January 2023. At this time, the first time range corresponding to the first time subordinate indicator being January is the same as the second time range corresponding to the second time subordinate indicator being January.

[0045] For another example, assuming that the first historical time range and the second historical time range are both 2021-2023, the subordinate indication duration of a first time subordinate indicator corresponding to a first time range is month, and the second time subordinate indicator corresponding to a corresponding second time range is the first quarter, the second time range corresponding to the second time subordinate indicator for the first quarter includes: the first time range corresponding to the first time subordinate indicator in January, the first time range corresponding to the first time subordinate indicator in February, and the first time range corresponding to the first time subordinate indicator in March, and so on.

[0046] S103, based on the observation data set, calculating the observation occurrence probability set within each first time range; and based on the forecast data set corresponding to each forecast scheme, calculating the forecast probability set of each forecast scheme within each first time range.

[0047] In an embodiment of the present invention, for any first time range in at least one first time range, the electronic device may calculate an observation occurrence probability set within any first time range (i.e., an observation occurrence probability set of the target area within any first time range) based on an observation event data set within any first time range (i.e., an observation event data set of the target area within any first time range).

[0048] Correspondingly, for any forecasting scheme among the N forecasting schemes, the electronic device can calculate the forecast probability set of any forecasting scheme within any first time range (that is, the forecast probability set of the target area within any first time range under any forecasting scheme) based on the forecast event data set of any forecasting scheme within any first time range (that is, the forecast event data set of the target area within any first time range under any forecasting scheme).

[0049] S104, based on the observed occurrence probability sets within each first time range, the forecast probability sets of each forecast scheme within each first time range, and the forecast conditional probability sets of each forecast scheme within each second time range, determine the occurrence probability of each target event in at least one target event under the event forecast result set, the event forecast result set including the event forecast results obtained by each forecast scheme for event forecasting the target area within the third time range.

[0050] Optionally, a target event can be regarded as an observed event. Optionally, at least one target event can include each observed event in multiple observed events, that is, each event in the event set, in which case each observed event can be regarded as a target event, that is, each event in the event set can be regarded as a target event; or, it can also include each event forecast result after deduplication in the event forecast result set (that is, each observed event that exists in multiple observed events and in the event forecast result set), in which case each event forecast result after deduplication in the event forecast result set can be regarded as a target event, and so on. Among them, an event forecast result can also be called an event forecast result of the target area within the third time range, and an event forecast result can be any forecast event among multiple forecast events.

[0051] Exemplarily, assuming that the value of N is 5, and the event forecast results obtained by each forecasting scheme for the target area within the third time range (that is, each forecasting scheme forecasts the event forecast results for the target area within the third time range) are event 1, event 1, event 1, event 3, and event 1, respectively. Then, when at least one target event includes each event forecast result after deduplication in the event forecast result set, at least one target event may include event 1 and event 3, and so on.

[0052] Optionally, the third time range may be any time range; optionally, the duration corresponding to the third time range may be one day (such as the third time range may be a day in the future), the duration corresponding to the third time range may be one month (such as the third time range may be a month in the future), and so on; the embodiment of the present invention is not limited to this.

[0053] The embodiment of the present invention can determine the forecast condition probability set of each forecast scheme in each second time range in at least one second time range based on the observation data set and the forecast data set corresponding to each forecast scheme after obtaining the observation data set and the forecast data set corresponding to each forecast scheme, respectively. The forecast condition probability set of a forecast scheme in a second time range is: the forecast condition probability set obtained by the corresponding forecast scheme for the target area in the corresponding second time range through event forecasting; the observation data set includes: the observation event data set of the target area in each first time range in at least one first time range, and a forecast data set includes: the forecast event data set of the target area in each first time range under the corresponding forecast scheme; wherein, a forecast event data is obtained by forecasting an event through the forecast scheme corresponding to the forecast data set, and N is a positive integer. Accordingly, the observation occurrence probability set in each first time range can be calculated based on the observation data set; and the forecast probability set of each forecast scheme in each first time range can be calculated based on the forecast data set corresponding to each forecast scheme. Based on this, the occurrence probability of each target event in at least one target event under the event forecast result set can be determined based on the observed occurrence probability set within each first time range, the forecast probability set of each forecast scheme within each first time range, and the forecast conditional probability set of each forecast scheme within each second time range. The event forecast result set includes the event forecast results obtained by each forecast scheme for forecasting the target area within the third time range. It can be seen that the embodiment of the present invention can specifically consider the forecast probability and forecast conditional probability under different forecast schemes through the observation data set and the forecast data set corresponding to each forecast scheme, that is, specifically consider the forecast probability and forecast conditional probability of each forecast event when different forecast schemes are used to forecast events, so as to accurately perform probability forecasting and obtain a probability forecasting result with higher accuracy (that is, the occurrence probability of each target event under the event forecast result set), which can effectively improve the accuracy of probability forecasting.

[0054] Based on the above description, the embodiment of the present invention further proposes a more specific probability forecasting method. Accordingly, the probability forecasting method can be executed by the electronic device (terminal or server) mentioned above; or, the probability forecasting method can be executed by the terminal and the server together. For the convenience of explanation, the following description is based on the example of an electronic device executing the probability forecasting method; please refer to Figure 2 , the probability forecasting method may include the following steps S201-S207:

[0055] S201, obtaining an observation data set and a forecast data set corresponding to each forecast scheme in N forecast schemes, wherein the observation data set includes: an observation event data set of a target area within each first time range in at least one first time range, and a forecast data set includes: a forecast event data set of the target area within each first time range under a corresponding forecast scheme; wherein one forecast event data is obtained by event forecasting through the forecast scheme corresponding to the forecast data set.

[0056] In an embodiment of the present invention, an observation event data set can be used to indicate the number of occurrences of each observation event among multiple observation events within the corresponding first time range, and a forecast event data set can be used to indicate the number of forecast events of each forecast event among multiple forecast events within the corresponding first time range of the corresponding forecast scheme.

[0057] It should be understood that when an observation event data is an observation event observed, or an observation event data is the number of occurrences of a certain observation event within a time sub-range within the corresponding first time range, an observation event data set may be an indication of the number of occurrences of each observation event within the corresponding first time range through the statistical results of the included observation event data; when an observation event data is the number of occurrences of a certain observation event within the corresponding first time range, an observation event data set may be an indication of the number of occurrences of each observation event within the corresponding first time range through the included observation event data. Similarly, when a forecast event data is a forecast event forecasted by a forecast scheme, or a forecast event data is the number of forecast events predicted by a forecast scheme within a time sub-range within a corresponding first time range, a forecast event data set may be an indication of the forecast number of each forecast event within the corresponding first time range through the statistical results of the included forecast event data; when a forecast event data is the number of forecast events predicted by a forecast scheme within the corresponding first time range for a forecast event, a forecast event data set may be an indication of the forecast number of each forecast event within the corresponding first time range through the included forecast event data, and so on.

[0058] S202, for any second time range in at least one second time range, determine P first time ranges corresponding to any second time range, and determine P observation event data sets within the first time range from the observation data set, where P is a positive integer.

[0059] For example, assuming that the first historical time range includes 2021-2023, the observation data set includes the observation data of the past three years, and the forecast data set corresponding to any forecast scheme includes the forecast data of the past three years, and assuming that the number of second time ranges in at least one second time range is 1, and the second time range is 2021-2023, the subordinate indicator duration of the first time subordinate indicator corresponding to a first time range is month, then the number of first time ranges in at least one first time range can be 12, and the second time range can correspond to 12 first time ranges (that is, the first time subordinate indicators are the first time ranges corresponding to each month), that is to say, the P first time ranges corresponding to any second time range can include each first time range, and the value of P can be 12.

[0060] For another example, assuming that both the first historical time range and the second historical time range include 2021-2023, the subordinate indicator duration of the first time subordinate indicator corresponding to a first time range is a month, and the subordinate indicator duration of the second time subordinate indicator corresponding to a second time range is a quarter, then a second time range may correspond to 3 first time ranges (such as the second time range corresponding to the first quarter of the second time subordinate indicator may include the first time range corresponding to each of the first time subordinate indicators in January, February and March), and at this time the value of P may be 3. It should be understood that the subordinate indicator duration of a second time subordinate indicator may be greater than or equal to the subordinate indicator duration of a first time subordinate indicator, then the P first time ranges corresponding to a second time range may include the first time range corresponding to each first time subordinate indicator included in the second time subordinate indicator of the corresponding second time range.

[0061] S203, traversing each forecasting scheme among the N forecasting schemes, taking the currently traversed forecasting scheme as the current forecasting scheme, and determining P forecast event data sets within the first time range from the forecast data sets corresponding to the current forecasting scheme.

[0062] S204, using P observation event data sets and forecast event data sets within the first time range, determine the forecast conditional probability set of the current forecast scheme within any second time range; wherein, a forecast conditional probability set includes a conditional occurrence probability set of a forecast scheme under the condition that each observation event among multiple observation events within a second time range has occurred, and a conditional occurrence probability set includes the occurrence probability of each forecast event among multiple forecast events under the condition that one observation event has occurred.

[0063] In the embodiment of the present invention, the probability of occurrence of a forecast event under the condition that an observed event has occurred can be a conditional occurrence probability. Among them, when the forecast event involved in a conditional occurrence probability set is the same as the observed event (such as the probability of occurrence of the forecast event "mild pollution" under the condition that the observed event "mild pollution" has occurred), the conditional occurrence probability can also be called the forecast accuracy, that is, under the condition that an observed event has occurred, the forecasted event is the probability of occurrence of the observed event.

[0064] In an embodiment of the present invention, the electronic device can traverse each observation event among multiple observation events, and use the currently traversed observation event as the current observation event, and use the observation event data set within P first time ranges (i.e., the P first time ranges corresponding to any second time range) to determine the number of occurrences of the current observation event; and can use the forecast event data set within the P first time ranges to respectively determine the forecast number of each forecast event under the condition that the current observation event has occurred. Based on this, the probability of occurrence of each forecast event under the condition that the current observation event has occurred can be calculated based on the number of occurrences of the current observation event and the forecast number of each forecast event under the condition that the current observation event has occurred, so as to obtain the conditional occurrence probability set of the current forecast scheme under the condition that the current observation event has occurred within any second time range; after traversing each observation event among multiple observation events, the forecast conditional probability set of the current forecast scheme within any second time range can be obtained.

[0065] Optionally, the electronic device may directly calculate the probability of occurrence of each predicted event under the condition that the current observed event has occurred, based on the number of occurrences of the current observed event and the number of forecasts of each predicted event under the condition that the current observed event has occurred. Exemplarily, assuming that the event set includes event 1, event 2 and event 3, the current observed event is event 1, and assuming that the number of occurrences of event 1 in P first time ranges corresponding to any second time range is 30, the number of forecasts of the forecast event "event 1" under the condition that the current observed event has occurred (at this time, it is the number of times the current forecast scheme correctly predicts the current observed event) is 15, the number of forecasts of the forecast event "event 2" under the condition that the current observed event has occurred (that is, the number of times the observation is the current observed event and forecast is event 2) is 7, and the number of forecasts of the forecast event "event 3" under the condition that the current observed event has occurred (that is, the number of times the observation is the current observed event and forecast is event 3) is 8; in this case, the probability of occurrence of the forecast event "event 1", the forecast event "event 2" and the forecast event "event 3" under the condition that the current observed event has occurred may be 15 / 30, 7 / 30 and 8 / 30, respectively.

[0066] Optionally, when calculating the probability of occurrence of each forecast event under the condition that the current observed event has occurred based on the number of occurrences of the current observed event and the forecast number of each forecast event under the condition that the current observed event has occurred, a conditional smoothing value can be determined; and the probability of occurrence of each forecast event under the condition that the current observed event has occurred is calculated based on the number of occurrences of the current observed event, the conditional smoothing value, and the forecast number of each forecast event under the condition that the current observed event has occurred. Optionally, the conditional smoothing value can be a constant greater than 0; optionally, the conditional smoothing value can be set according to experience or actual needs, and the embodiment of the present invention does not limit this; illustratively, a smaller value can be taken when the sample size is small, such as when the conditional smoothing value can be 0.1, etc. Based on this, the conditional smoothing value can be used to achieve smoothing processing (such as Laplace smoothing processing) of the probability of occurrence of each forecast event under the condition that the current observed event has occurred.

[0067] Optionally, for any forecast event among the multiple forecast events, the electronic device may use formula 1.1 to calculate the probability of occurrence of any forecast event under the condition that the current observed event has occurred:

[0068] Formula 1.1

[0069] Among them, x n can represent any forecast event predicted by the nth forecast scheme (that is, any forecast event predicted when the current forecast scheme is the nth forecast scheme), y can represent any observed event (here, the current observed event), M y It can represent the number of occurrences of the current observation event, M xn It can represent the number of forecasts of any forecast event under the condition that the current observed event has occurred (that is, the number of forecasts of any forecast event predicted by the nth forecast scheme under the condition that the current observed event has occurred), and α can represent the conditional smoothing value; wherein, n∈[1, N], and n is an integer. Based on this, the embodiment of the present invention can avoid the situation where the probability is 0. In other words, the probability of occurrence of any forecast event under the condition that the current observed event has occurred can be: (the number of forecasts of any forecast event under the condition that the current observed event has occurred + conditional smoothing value) / (the number of occurrences of the current observed event + conditional smoothing value × N).

[0070] Optionally, the electronic device may also determine whether there is a forecast number less than the conditional smoothing number threshold in the forecast number of each forecast event under the condition that the current observed event has occurred; if there is a forecast number less than the conditional smoothing number threshold in the forecast number of each forecast event under the condition that the current observed event has occurred, the above-mentioned determination of the conditional smoothing value is triggered; if there is no forecast number less than the conditional smoothing number threshold in the forecast number of each forecast event under the condition that the current observed event has occurred, the above-mentioned direct calculation of the probability of occurrence of each forecast event under the condition that the current observed event has occurred based on the number of occurrences of the current observed event and the forecast number of each forecast event under the condition that the current observed event has occurred is triggered. Optionally, the conditional smoothing number threshold may be set according to experience or actual needs, and the embodiment of the present invention does not limit this; for example, the conditional smoothing number threshold may be 1, and at this time, when there is a zero value in the forecast number of each forecast event under the condition that the current observed event has occurred, it can be determined that there is a forecast number less than the conditional smoothing number threshold.

[0071] In one embodiment, one observation event data is one observation event, one forecast event data is one forecast event forecasted by a forecasting scheme, and the observation event data in the P observation event data sets within the first time range correspond one to one with the forecast event data in the P forecast event data sets within the first time range. Based on this, when using the P observation event data sets within the first time range to determine the number of occurrences of the current observation event, the electronic device can perform data statistics on the current observation event in the P observation event data sets within the first time range to obtain the number of occurrences of the current observation event. Correspondingly, when using P forecast event data sets within the first time range to determine the forecast times of each forecast event under the condition that the current observation event has occurred, for any forecast event among the multiple forecast events, the electronic device can traverse the forecast event data corresponding to the current observation event in the P forecast event data sets within the first time range (that is, the observation event data is the forecast event data corresponding to each observation event data of the current observation event), and use the currently traversed forecast event data as the current forecast event data; if the current forecast event data is any of the forecast events, the forecast times of any forecast event under the condition that the current observation event has occurred is accumulated to update the forecast times of any forecast event under the condition that the current observation event has occurred, so that the forecast times of any forecast event under the condition that the current observation event has occurred is increased by 1, and the initial state of the forecast times of any forecast event under the condition that the current observation event has occurred can be 0; after traversing the forecast event data sets corresponding to the current observation event in the P forecast event data sets within the first time range, the forecast times of any forecast event under the condition that the current observation event has occurred can be obtained.

[0072] In another implementation, a forecast event data may include a forecast event forecasted by a forecast scheme and an observation identifier, and an observation identifier may be used to indicate the observation event corresponding to the corresponding forecast event data. Based on this, when using P forecast event data sets within the first time range to respectively determine the forecast times of each forecast event under the condition that the current observation event has occurred, for any forecast event among the multiple forecast events (such as mild pollution), the forecast event data may be the forecast event data of any forecast event, and the observation identifier may be the observation identifier of the current observation event. Data statistics may be performed to obtain the forecast times of any forecast event under the condition that the current observation event has occurred. Optionally, the observation identifier of an observation event may be set according to experience or actual needs, and the embodiment of the present invention is not limited to this.

[0073] In another implementation, a forecast event data may be the forecast number of a forecast event under the condition that an observed event has occurred. Based on this, the electronic device can determine the forecast number of any forecast event in each first time range under the condition that the current observed event has occurred from the forecast event data set in each first time range in the P first time ranges, and sum the forecast number of any forecast event in each first time range under the condition that the current observed event has occurred, to obtain the forecast number of any forecast event under the condition that the current observed event has occurred, and so on.

[0074] S205, after traversing each forecasting scheme in the N forecasting schemes, obtain a forecast conditional probability set of each forecasting scheme in any second time range. The forecast conditional probability set of a forecasting scheme in a second time range is: the forecast conditional probability set obtained by the corresponding forecasting scheme for event forecasting of the target area in the corresponding second time range.

[0075] S206, calculating the observation occurrence probability set within each first time range based on the observation data set; and calculating the forecast probability set of each forecast scheme within each first time range based on the forecast data set corresponding to each forecast scheme.

[0076] In an embodiment of the present invention, when calculating the observed occurrence probability set within each first time range based on the observation data set, the electronic device may calculate the observed occurrence probability set within each first time range based on the number of occurrences indicated by the observed event data set within each first time range, and one observed occurrence probability set may include the observed occurrence probability of each observed event within the corresponding first time range. Correspondingly, when calculating the forecast probability set of each forecast scheme within each first time range based on the forecast data set corresponding to each forecast scheme, for any forecast scheme among the N forecast schemes, the electronic device may calculate the forecast probability set of any forecast scheme within each first time range based on the number of forecasts indicated by the forecast event data set within each first time range, and one forecast probability set may include the forecast probability of each forecast event within the corresponding first time range of the corresponding forecast scheme. Among them, the number of occurrences indicated by the observed event data set within a first time range may include the number of occurrences of each observed event within the corresponding first time range, and the number of forecasts indicated by the forecast event data set within a first time range of any forecast scheme may include the forecast number of forecast events of any forecast scheme within the corresponding first time range.

[0077] In one embodiment, for any first time range in at least one first time range, the electronic device may determine an observation smoothing value, and calculate an observation occurrence probability set in any first time range based on the observation smoothing value and the number of occurrences indicated by the observation event data set in any first time range; in this case, the number of occurrences indicated by the observation event data set in any first time range may be summed to obtain the total number of occurrences of the observation events in the first time range, so that the observation occurrence probability set in any first time range may be calculated using the observation smoothing value, the number of occurrences indicated by the observation event data set in any first time range, and the total number of occurrences of the observation events in any first time range. Based on this, for any observation event in multiple observation events, the observation smoothing value, the number of occurrences of any observation event in any first time range, and the total number of occurrences of the observation events in any first time range may be used to calculate the observation occurrence probability of any observation event in any first time range; illustratively, at this time, the observation occurrence probability of any observation event in any first time range may be: (number of occurrences of any observation event in any first time range + observation smoothing value) / (total number of occurrences of observation events in any first time range + observation smoothing value × N). Optionally, the observation smoothing value may be set according to experience or according to actual needs, which is not limited in the embodiment of the present invention.

[0078] Optionally, the electronic device may also determine a forecast smoothing value, and calculate a forecast probability set for any forecast scheme within any first time range based on the forecast smoothing value and the forecast number of times indicated by the forecast event data set for any forecast scheme within any first time range; in this case, the forecast number of times indicated by the forecast event data set for any forecast scheme within any first time range may be summed to obtain the total number of event forecasts for any forecast scheme within any first time range, thereby calculating the forecast probability set for any forecast scheme within any first time range using the forecast smoothing value, the forecast number of times indicated by the forecast event data set for any forecast scheme within any first time range, and the total number of event forecasts for any forecast scheme within any first time range. Based on this, for any forecast event among the multiple forecast events, the forecast smoothing value, the forecast times of any forecast event in any forecast scheme in any first time range, and the sum of the forecast times of events in any forecast scheme in any first time range can be used to calculate the forecast probability of any forecast event in any forecast scheme in any first time range; illustratively, at this time, the forecast probability of any forecast event in any forecast scheme in any first time range can be: (the forecast times of any forecast event in any forecast scheme in any first time range + forecast smoothing value) / (the sum of the forecast times of events in any forecast scheme in any first time range + forecast smoothing value × N). Optionally, the forecast smoothing value can be set according to experience or according to actual needs, and the embodiment of the present invention is not limited to this.

[0079] Optionally, any two smoothed values ​​among the conditional smoothed values, the observed smoothed values, and the predicted smoothed values ​​may be the same or different, and the embodiment of the present invention does not limit this. Based on this, the smoothing processing of the observed occurrence probability set within any first time range can be realized, and the smoothing processing of the forecast probability set of any forecast scheme within any first time range can be realized, which can make the observed occurrence probability and the forecast probability more reasonable to avoid the occurrence of probabilities such as 0 or 1.

[0080] In another embodiment, for any first time range in at least one first time range, the electronic device may sum the number of occurrences indicated by the observation event data set in any first time range to obtain the total number of occurrences of the observation events in the any first time range, so that the number of occurrences indicated by the observation event data set in any first time range and the total number of occurrences of the observation events in any first time range may be used to calculate the observation occurrence probability set in any first time range. Based on this, for any observation event in multiple observation events, the number of occurrences of any observation event in any first time range and the total number of occurrences of the observation events in any first time range may be used to calculate the observation occurrence probability of any observation event in any first time range; illustratively, at this time, the observation occurrence probability of any observation event in any first time range may be: the number of occurrences of any observation event in any first time range / the total number of occurrences of the observation events in any first time range.

[0081] Optionally, the electronic device may also sum the forecast times indicated by the forecast event data set of any forecast scheme within any first time range to obtain the sum of the forecast times of events of any forecast scheme within any first time range, so as to calculate the forecast probability set of any forecast scheme within any first time range by using the forecast times indicated by the forecast event data set of any forecast scheme within any first time range and the sum of the forecast times of events of any forecast scheme within any first time range. Based on this, for any forecast event among multiple forecast events, the forecast times of any forecast event of any forecast scheme within any first time range and the sum of the forecast times of events of any forecast scheme within any first time range may be used to calculate the forecast probability of any forecast event of any forecast scheme within any first time range; illustratively, at this time, the forecast probability of any forecast event of any forecast scheme within any first time range may be: the forecast times of any forecast event of any forecast scheme within any first time range / the sum of the forecast times of events of any forecast scheme within any first time range.

[0082] In another embodiment, the electronic device may also determine whether there is a number of occurrences less than the observation smoothing number threshold value among the number of occurrences indicated by the observation event data set within any first time range; if there is a number of occurrences less than the observation smoothing number threshold value among the number of occurrences indicated by the observation event data set within any first time range, the above determination of the observation smoothing value is triggered; if there is no number of occurrences less than the observation smoothing number threshold value among the number of occurrences indicated by the observation event data set within any first time range, the above summing processing of the number of occurrences indicated by the observation event data set within any first time range is triggered to obtain the total number of occurrences of the observation events within any first time range, so that the number of occurrences indicated by the observation event data set within any first time range and the total number of occurrences of the observation events within any first time range can be used to calculate the observation occurrence probability set within any first time range. Optionally, the observation smoothing number threshold value may be set according to experience or according to actual needs, and the embodiment of the present invention is not limited to this.

[0083] Correspondingly, the electronic device can also determine whether there is a forecast number less than the forecast smoothing number threshold value among the forecast times indicated by the forecast event data set of any forecast scheme within any first time range; if there is a forecast number less than the forecast smoothing number threshold value among the forecast times indicated by the forecast event data set of any forecast scheme within any first time range, then trigger the execution of the above-mentioned determination of the forecast smoothing value; if there is no forecast number less than the forecast smoothing number threshold value among the forecast times indicated by the forecast event data set of any forecast scheme within any first time range, then trigger the execution of the above-mentioned use of the forecast times indicated by the forecast event data set of any forecast scheme within any first time range and the sum of the event forecast times of any forecast scheme within any first time range, and calculate the forecast probability set of any forecast scheme within any first time range. Optionally, the forecast smoothing number threshold value can be set according to experience or according to actual needs, and the embodiment of the present invention does not limit this.

[0084] S207, based on the observed occurrence probability set within each first time range, the forecast probability set of each forecast scheme within each first time range, and the forecast conditional probability set of each forecast scheme within each second time range, determine the occurrence probability of each target event in at least one target event under the event forecast result set, and the event forecast result set includes the event forecast results obtained by each forecast scheme for event forecasting the target area within the third time range.

[0085] In an embodiment of the present invention, the electronic device may respectively perform event forecasting for the target area within the third time range through various forecasting schemes, thereby obtaining event forecasting results obtained by each forecasting scheme respectively performing event forecasting for the target area within the third time range. The event forecasting result obtained by a forecasting scheme performing event forecasting for the target area within the third time range may also be referred to as the event forecasting result of the corresponding forecasting scheme for the target area within the third time range.

[0086] Based on this, the electronic device can iteratively update the posterior probability (i.e., the probability of occurrence of each target event under the event forecast result set) based on the prior probability (i.e., the set of observed occurrence probabilities within each first time range) and the event forecast result set through the Naive Bayes method (Naive Bayes). Among them, the Naive Bayes method is a set of supervised learning algorithms based on Bayes' theorem. Given the value of the class variable (i.e., any target event, which can be an observed event), it is assumed that the conditional independence between each pair of features (i.e., the event forecast result set) is a "naive" assumption. In an embodiment of the present invention, formula 1.2 can be used to represent the class variable y (such as corresponding to different observed events) and the event forecast result set x1 to x N (x here n It can represent the forecast result of the event predicted by the nth forecast scheme, and can also be called the nth event forecast result in the event forecast result set, that is, it represents a forecast event) The relationship between:

[0087] Formula 1.2

[0088] Among them, P(y|x1,…,x N ) can represent the probability of occurrence of any target event y under the event forecast result set (i.e., the posterior probability of occurrence of any target event under the event forecast result set), p(y) can represent the probability of occurrence of the target observation under any target event in the third time range (i.e., the probability of occurrence of the observed event in the first time range matching the third time range being any target event), P(x1,…,x N |y) can represent the probability of occurrence of the event forecast result set under the condition that any target event has occurred (that is, the probability of occurrence of the event forecast result set under the condition that the observed event is any target event). N ) can represent the forecast probability of the event forecast result set (that is, the probability of forecasting the event forecast result set).

[0089] Exemplarily, assuming that the value of N is 3, and the event forecast results of the three forecast schemes for the target area in the third time range (such as tomorrow) are light pollution, light pollution and moderate pollution respectively, then x1 (i.e., the event forecast result predicted by the first forecast scheme, also referred to as the first event forecast result) can be light pollution, x2 (i.e., the event forecast result predicted by the second forecast scheme) can be light pollution, and x3 (i.e., the event forecast result predicted by the third forecast scheme) can be moderate pollution; assuming that any target event is light pollution, then the probability of light pollution occurring in the target area in the third time range under the combination of N forecast schemes can be calculated, that is, y can be light pollution at this time, then P(y|x1,…,x N ) can represent the posterior probability that the target area will experience light pollution in the third time range when the event forecast results of the three forecast schemes for the target area in the third time range are light pollution, light pollution, and moderate pollution respectively. That is to say, at this time, the event forecast result set can include light pollution, light pollution, and moderate pollution respectively, thereby representing the probability of occurrence of any target event "light pollution" in the event forecast result set; P(x1,…,x N |y) can represent the probability that the event forecast results of the target area of ​​the three forecast schemes in the third time range are respectively light pollution, light pollution and moderate pollution when the observed event is any target event "light pollution", that is, it represents the probability of occurrence of the event forecast result set under the condition that any target event "light pollution" has occurred; P(x1,…,x N ) can represent the forecast probability of the combination of light pollution, mild pollution and moderate pollution in the third time range for the event forecast results of the target area of ​​the three forecast schemes in all forecast samples.

[0090] In the embodiment of the present invention, a simple conditional independence assumption can be used, that is, first assuming that the event forecast results predicted by each forecasting scheme are independent of each other, which can be shown as formula 1.3:

[0091] Formula 1.3

[0092] Among them, P(x n ) can represent the forecast probability of the event forecast result of the target area in the third time range predicted by the nth forecast scheme (i.e., the forecast probability of the event forecast result obtained by the nth forecast scheme for the target area in the third time range), such as x n When x is slightly polluted, it can represent the forecast probability that the event forecast result of the target area of ​​the nth forecast scheme in the third time range is slightly polluted; correspondingly, when x n When any target event is light pollution, P(x n|y) can represent the probability of occurrence of the event forecast result "mild pollution" under the condition that any target event "mild pollution" has occurred, that is, when the actual pollution level is light pollution, the probability that the event forecast result predicted by the nth forecast scheme is light pollution.

[0093] In summary, it can be seen that the embodiments of the present invention can calculate the probability of occurrence of any target event under the event forecast result set through the forecast probability of the third time range under each event forecast result (the forecast probability of a third time range under an event forecast result may refer to the forecast probability of the event forecast result obtained by the corresponding forecast scheme for the target area under the third time range for event forecasting), the occurrence probability of each event forecast result under the condition that any target event has occurred, and the occurrence probability of the target observation under any target event in the third time range.

[0094] Accordingly, for any target event among at least one target event, the electronic device can determine a first time range that matches the third time range from at least one first time range, and select the target observation occurrence probability of the third time range under any target event from the set of observation occurrence probabilities within the determined first time range.

[0095] Specifically, when determining a first time range that matches the third time range from at least one first time range, the first time subordinate indicators corresponding to each first time range and the first time subordinate indicator corresponding to the third time range can be determined; and based on the first time subordinate indicators corresponding to each first time range and the first time subordinate indicator corresponding to the third time range, the first time range that matches the third time range is determined from at least one first time range, and the determined first time range (i.e., the first time range that matches the third time range) is: the first time range whose first time subordinate indicator is the same as the first time subordinate indicator corresponding to the third time range. Exemplarily, taking the subordinate indicator duration of the first time subordinate indicator as a month as an example, assuming that the third time range is July 30, 2024, the electronic device can determine that the first time subordinate indicator corresponding to the third time range is July, and the determined first time range is: the first time range corresponding to the first time subordinate indicator in July.

[0096] Optionally, when selecting the target observation occurrence probability of the third time range under any target event from the determined observation occurrence probability set within the first time range, the observed occurrence probability of the determined observed event within the first time range being any target event can be selected from the determined observation occurrence probability set within the first time range, and the observed occurrence probability of the determined observed event within the first time range being any target event is used as the target observation occurrence probability under any target event in the third time range. Exemplarily, assuming that any target event is light pollution, the target observation occurrence probability of the third time range under any target event can be the observed occurrence probability of the observed event "light pollution" within the determined first time range.

[0097] Furthermore, the electronic device selects the forecast probabilities of each event forecast result in the third time range in the event forecast result set from the forecast probability set of each forecast scheme in the determined first time range. The forecast probability of the third time range under an event forecast result may be: the forecast probability of the forecast scheme that forecasts the corresponding event forecast result within the determined first time range that the forecast event is the corresponding event forecast result. For example, when an event forecast result is light pollution, the forecast probability of the third time range under the event forecast result "light pollution" is the forecast probability of the forecast scheme that forecasts the event forecast result "light pollution" within the determined first time range that forecasts the event "light pollution".

[0098] Correspondingly, the electronic device can determine a second time range that matches the third time range from at least one second time range, and select the probability of occurrence of each event forecast result under the condition that any target event has occurred from the forecast condition probability set of each forecast scheme within the determined second time range.

[0099] Optionally, when determining a second time range that matches the third time range from at least one second time range, the second time subordinate indicator corresponding to each second time range and the second time subordinate indicator corresponding to the third time range can be determined; and based on the second time subordinate indicators corresponding to each second time range and the second time subordinate indicator corresponding to the third time range, the second time range that matches the third time range is determined from at least one second time range, and the determined second time range (i.e., the second time range that matches the third time range) can be: a second time range whose second time subordinate indicator is the same as the second time subordinate indicator corresponding to the third time range. Exemplarily, taking the subordinate indicator duration of the second time subordinate indicator as a quarter as an example, assuming that the third time range is July 30, 2024, the electronic device can determine that the second time subordinate indicator corresponding to the third time range is the third quarter, and the determined second time range can be: a second time range corresponding to the second time subordinate indicator being the third quarter.

[0100] Optionally, when selecting the probability of occurrence of each event forecast result under the condition that any target event has occurred from the forecast condition probability set of each forecast scheme within the determined second time range, for any forecast scheme among the N forecast schemes, the probability of occurrence of the event forecast result predicted by any forecast scheme under the condition that any target event has occurred can be selected from the forecast condition probability set of any forecast scheme within the determined second time range; that is, the probability of occurrence of the event forecast result predicted by any forecast scheme within the determined second time range as the event forecast result predicted by any forecast scheme under the condition that any target event (that is, the observed event at this time is any target event) has occurred can be selected from the forecast condition probability set of any forecast scheme within the determined second time range as the occurrence probability of the event forecast result predicted by any forecast scheme under the condition that any target event has occurred. For example, when the event forecast result predicted by any forecast scheme is moderate pollution and any target event is light pollution, the probability of occurrence of the event forecast result predicted by any forecast scheme under the condition that any target event has occurred is: the probability of occurrence of the forecast event "moderate pollution" predicted by any forecast scheme within the determined second time range under the condition that the observed event "light pollution" has occurred.

[0101] Based on this, the electronic device can calculate the probability of occurrence of any target event in the event forecast result set based on the target observation occurrence probability, the forecast probability of the third time range under each event forecast result, and the probability of occurrence of each event forecast result under the condition that any target event has occurred. In an embodiment of the present invention, each event forecast result in the event forecast result set can be traversed in sequence, and the currently traversed event forecast result can be used as the current event forecast result; when the current event forecast result is the first event forecast result in the event forecast result set, the probability of occurrence of any target event in the result set corresponding to the current event forecast result can be calculated based on the target observation occurrence probability, the forecast probability of the third time range under the current event forecast result, and the probability of occurrence of the current event forecast result under the condition that any target event has occurred; when the current event forecast result is not the first event forecast result, the result set corresponding to the previous event forecast result of any target event in the current event forecast result can be used. The probability of occurrence under the third time range, the forecast probability under the current event forecast result, and the probability of occurrence of the current event forecast result under the condition that any target event has occurred, calculate the probability of occurrence of any target event under the result set corresponding to the current event forecast result; after traversing each event forecast result in the event forecast result set, the probability of occurrence of any target event under the event forecast result set is obtained, and the probability of occurrence of any target event under the event forecast result set is the probability of occurrence of any target event under the result set corresponding to the Nth event forecast result in the event forecast result set. The result set corresponding to an event forecast result includes each event forecast result from the first event forecast result to the corresponding event forecast result. Among them, the event forecast result set may include N event forecast results, and an event forecast result is obtained by a forecast scheme for event forecasting the target area under the third time range. For example, assuming that the current event forecast result is the nth event forecast result, then the result set corresponding to the current event forecast result may include the first event forecast result to the nth event forecast result (i.e., x1,…,x n ).

[0102] Optionally, taking the current event forecast result as the nth event forecast result as an example, the electronic device can use formula 1.4 to calculate the probability of occurrence of any target event under the result set corresponding to the current event forecast result:

[0103] Formula 1.4

[0104] Among them, P(y|x1,…,x n ) can represent the probability of occurrence of any target event under the result set corresponding to the current event forecast result; when n is equal to N, the probability of occurrence of any target event under the event forecast result set can be obtained. Correspondingly, when n is not 1, P(y|x1,…,xn-1 ) can represent the probability of occurrence of any target event under the result set corresponding to the previous event forecast result of the current event forecast result; when n is 1, P(y|x1,…,x n-1 ) can be P(y) (i.e., the probability of occurrence of the target observation).

[0105] In summary, the embodiment of the present invention can traverse each event forecast result in the event forecast result set, that is, continuously update the event forecast result predicted for the target area in the third time range according to the nth forecast scheme in the feature combination x1, ..., x n The probability of any target event occurring at this time is P(y|x1,…,x n ), until we get P(y|x1,…,x N ). Based on this, the embodiment of the present invention can convert the influence of the event forecast results of N forecast schemes on the occurrence probability into Bayesian iterative update through the naive conditional independence assumption, which can greatly simplify the difficulty of engineering implementation, thereby quickly obtaining the occurrence probability of any target event under the event forecast result set.

[0106] After acquiring the observation data set and the forecast data set corresponding to each of N forecast schemes, the embodiment of the present invention can determine, for any second time range in at least one second time range, the P first time ranges corresponding to any second time range, and determine the observation event data sets within the P first time ranges from the observation data set. Then, each forecast scheme in the N forecast schemes can be traversed, and the currently traversed forecast scheme can be used as the current forecast scheme, and P forecast event data sets within the first time range can be determined from the forecast data set corresponding to the current forecast scheme; and the P observation event data sets and forecast event data sets within the first time range can be used to determine the forecast conditional probability set of the current forecast scheme within any second time range; wherein, a forecast conditional probability set includes a conditional occurrence probability set under the condition that each observation event among multiple observation events within a second time range has occurred in a forecast scheme, and a conditional occurrence probability set includes the occurrence probability of each forecast event among multiple forecast events under the condition that one observation event has occurred; after traversing each forecast scheme in the N forecast schemes, the forecast conditional probability set of each forecast scheme within any second time range is obtained, and the forecast conditional probability set of a forecast scheme within a second time range is: the forecast conditional probability set obtained by the corresponding forecast scheme for event forecasting of the target area within the corresponding second time range. Accordingly, based on the observation data set, the observation occurrence probability set within each first time range can be calculated; and based on the forecast data set corresponding to each forecast scheme, the forecast probability set of each forecast scheme within each first time range can be calculated. Based on this, based on the observation occurrence probability set within each first time range, the forecast probability set of each forecast scheme within each first time range, and the forecast condition probability set of each forecast scheme within each second time range, the occurrence probability of each target event in at least one target event under the event forecast result set can be determined, and the event forecast result set includes the event forecast results obtained by each forecast scheme for the target area under the third time range. It can be seen that the embodiment of the present invention can analyze each forecast scheme in a targeted manner, that is, the embodiment of the present invention adopts the probability forecast method of the naive Bayes method considering the prior probability, which can more reasonably forecast the uncertainty (i.e., forecast occurrence probability) under the condition of comprehensively considering the forecasts of N forecast schemes, thereby obtaining the corresponding posterior probability forecast, which can effectively improve the accuracy of the probability forecast. Moreover, it can reduce the dependence on the forecast effect of a single forecast scheme, prevent the impact of the poor forecast accuracy of the forecast scheme when only referring to one forecast scheme, and effectively avoid the situation where the forecast results of a certain forecast scheme for certain events have no reference value at all. It can effectively avoid the inability of the posterior probability to provide valuable reference, and at the same time, it can make full use of the available event forecast results to comprehensively obtain a more reasonable probability forecast.In addition, the embodiment of the present invention can avoid probabilities of 0 or 1, etc. through smoothing processing. This is because the probability can be very low but future events cannot be confirmed. The embodiment of the present invention can make the corresponding probability more reasonable; that is, the embodiment of the present invention can effectively solve the defect that an unreasonable probability of 0 or 1 may be given when implementing probability forecasts for event forecast results based on a variety of different forecasting schemes. The embodiment of the present invention can introduce a priori probability and smoothed maximum likelihood estimation, and iteratively update the output of a more reasonable posterior probability based on the naive Bayes model on the basis of the prior probability, so as to further improve the accuracy of the probability forecast.

[0107] Based on the description of the related embodiments of the above-mentioned probability forecasting method, the embodiment of the present invention further proposes a probability forecasting device, which can be a computer program (including program code) running in an electronic device; Figure 3 As shown, the probability forecasting device may include an acquisition unit 301 and a processing unit 302. The probability forecasting device may execute Figure 1 or Figure 2 The probability forecasting method shown, that is, the probability forecasting device can run the above units:

[0108] The acquisition unit 301 is used to acquire an observation data set and a forecast data set corresponding to each forecast scheme in N forecast schemes, wherein the observation data set includes: an observation event data set of a target area in each first time range in at least one first time range, and a forecast data set includes: a forecast event data set of the target area in each first time range under a corresponding forecast scheme; wherein one forecast event data is obtained by forecasting an event through the forecast scheme corresponding to the forecast data set, and N is a positive integer;

[0109] The processing unit 302 is used to determine, based on the observation data set and the forecast data set corresponding to each forecast scheme, a forecast condition probability set of each forecast scheme in each second time range in at least one second time range, wherein the forecast condition probability set of a forecast scheme in a second time range is: the forecast condition probability set obtained by the corresponding forecast scheme performing event forecasting on the target area in the corresponding second time range;

[0110] The processing unit 302 is further configured to calculate the observation occurrence probability set within each first time range based on the observation data set; and calculate the forecast probability set of each forecast scheme within each first time range based on the forecast data set corresponding to each forecast scheme;

[0111] The processing unit 302 is also used to determine the occurrence probability of each target event in at least one target event under an event forecast result set based on the observed occurrence probability set within the each first time range, the forecast probability set of each forecast scheme within the each first time range, and the forecast conditional probability set of each forecast scheme within the each second time range. The event forecast result set includes the event forecast results obtained by each forecast scheme for forecasting the target area within the third time range.

[0112] In one embodiment, when the processing unit 302 determines the forecast conditional probability set of each forecast scheme in each second time range in at least one second time range based on the observation data set and the forecast data set corresponding to each forecast scheme, it can be specifically used to: determine, for any second time range in the at least one second time range, P first time ranges corresponding to the any second time range, and determine the observation event data set within the P first time ranges from the observation data set, where P is a positive integer; traverse each forecast scheme in the N forecast schemes, and use the currently traversed forecast scheme as the current forecast scheme, and determine the observation event data set from the forecast data set corresponding to the current forecast scheme. Determine the forecast event data sets within the P first time ranges; use the observed event data sets and the forecast event data sets within the P first time ranges to determine the forecast conditional probability set of the current forecast scheme within any of the second time ranges; wherein, a forecast conditional probability set includes a conditional occurrence probability set of a forecast scheme under the condition that each observed event among multiple observed events within a second time range has occurred, and a conditional occurrence probability set includes the occurrence probability of each forecast event among multiple forecast events under the condition that one observed event has occurred; after traversing each of the N forecast schemes, obtain the forecast conditional probability set of each forecast scheme within any of the second time ranges.

[0113] In another embodiment, when the processing unit 302 uses the observed event data sets and the forecast event data sets within the P first time ranges to determine the forecast conditional probability set of the current forecast scheme within the second time range, it can be specifically used to: traverse each observed event among the multiple observed events, and take the currently traversed observed event as the current observed event, and use the P observed event data sets within the first time range to determine the number of occurrences of the current observed event; use the forecast event data sets within the P first time ranges to respectively determine the forecast number of each forecast event under the condition that the current observed event has occurred; calculate the occurrence probability of each forecast event under the condition that the current observed event has occurred based on the number of occurrences of the current observed event and the forecast number of each forecast event under the condition that the current observed event has occurred, so as to obtain the conditional occurrence probability set of the current forecast scheme under the condition that the current observed event has occurred within the second time range; after traversing each observed event among the multiple observed events, obtain the forecast conditional probability set of the current forecast scheme within the second time range.

[0114] In another embodiment, when the processing unit 302 calculates the probability of occurrence of each forecast event under the condition that the current observed event has occurred based on the number of occurrences of the current observed event and the forecast number of each forecast event under the condition that the current observed event has occurred, it can be specifically used to: determine a conditional smoothing value; calculate the probability of occurrence of each forecast event under the condition that the current observed event has occurred based on the number of occurrences of the current observed event, the conditional smoothing value and the forecast number of each forecast event under the condition that the current observed event has occurred.

[0115] In another implementation, an observed event data set is used to indicate the number of occurrences of each observed event among multiple observed events within the corresponding first time range, and a forecast event data set is used to indicate the forecast number of each forecast event among multiple forecast events within the corresponding first time range of the corresponding forecast scheme; when the processing unit 302 calculates the observed occurrence probability sets within the respective first time ranges based on the observed data set, it can be specifically used to: calculate the observed occurrence probability sets within the respective first time ranges based on the number of occurrences indicated by the observed event data sets within the respective first time ranges, and one observed occurrence probability set includes the observed occurrence probabilities of the respective observed events within the corresponding first time range;

[0116] When the processing unit 302 calculates the forecast probability sets of each forecast scheme within the first time range based on the forecast data sets corresponding to each forecast scheme, it can be specifically used to: for any forecast scheme among the N forecast schemes, based on the number of forecasts indicated by the forecast event data sets of any forecast scheme within the first time range, calculate the forecast probability set of any forecast scheme within the first time range, one forecast probability set including the forecast probability of each forecast event within the corresponding first time range of the corresponding forecast scheme.

[0117] In another implementation manner, when the processing unit 302 calculates the observed occurrence probability set within each first time range based on the number of occurrences indicated by the observed event data set within each first time range, the processing unit 302 may be specifically configured to: determine an observed smoothing value for any first time range of the at least one first time range, and calculate the observed occurrence probability set within any first time range based on the observed smoothing value and the number of occurrences indicated by the observed event data set within any first time range;

[0118] When the processing unit 302 calculates the forecast probability set of any forecast scheme within each first time range based on the forecast number of times indicated by the forecast event data set of any forecast scheme within each first time range, it can be specifically used to: determine the forecast smoothing value, and calculate the forecast probability set of any forecast scheme within any first time range based on the forecast smoothing value and the forecast number of times indicated by the forecast event data set of any forecast scheme within the first time range.

[0119] In another embodiment, when the processing unit 302 determines the occurrence probability of each target event in at least one target event under the event forecast result set based on the observed occurrence probability set within the each first time range, the forecast probability set of the each forecast scheme within the each first time range, and the forecast condition probability set of the each forecast scheme within the each second time range, it can be specifically used to: for any target event among the at least one target event, determine a first time range matching the third time range from the at least one first time range, and select the target observation occurrence probability of the third time range under the any target event from the determined observed occurrence probability set within the first time range; respectively select the target observation occurrence probability of the third time range from the each forecast scheme in the From the forecast probability set within the determined first time range, the forecast probability of the third time range under each event forecast result in the event forecast result set is selected; from the at least one second time range, a second time range matching the third time range is determined, and from the forecast conditional probability set of each forecast scheme within the determined second time range, the occurrence probability of each event forecast result under the condition that any one of the target events has occurred is selected; based on the target observation occurrence probability, the forecast probability of the third time range under each event forecast result and the occurrence probability of each event forecast result under the condition that any one of the target events has occurred, the occurrence probability of any one of the target events under the event forecast result set is calculated.

[0120] In another embodiment, when the processing unit 302 calculates the probability of occurrence of any target event in the event forecast result set based on the target observation occurrence probability, the forecast probability of the third time range under the respective event forecast results, and the occurrence probability of the respective event forecast results under the condition that any target event has occurred, it can be specifically used to: traverse each event forecast result in the event forecast result set in turn, and use the currently traversed event forecast result as the current event forecast result; when the current event forecast result is the first event forecast result in the event forecast result set, calculate the probability of occurrence of any target event in the result set corresponding to the current event forecast result based on the target observation occurrence probability, the forecast probability of the third time range under the current event forecast result, and the occurrence probability of the current event forecast result under the condition that any target event has occurred; when the current event forecast result is the first event forecast result in the event forecast result set, calculate the probability of occurrence of any target event in the result set corresponding to the current event forecast result based on the target observation occurrence probability, the forecast probability of the third time range under the current event forecast result, and the occurrence probability of the current event forecast result under the condition that any target event has occurred; When the result is not the first event forecast result, the probability of occurrence of any target event in the result set corresponding to the current event forecast result is calculated based on the probability of occurrence of any target event in the result set corresponding to the previous event forecast result of the current event forecast result, the forecast probability of the third time range under the current event forecast result, and the probability of occurrence of the current event forecast result under the condition that any target event has occurred; after traversing each event forecast result in the event forecast result set, the probability of occurrence of any target event in the event forecast result set is obtained, and the probability of occurrence of any target event in the event forecast result set is the probability of occurrence of any target event in the result set corresponding to the Nth event forecast result in the event forecast result set, and the result set corresponding to one event forecast result includes each event forecast result from the first event forecast result to the corresponding event forecast result.

[0121] In another implementation manner, when the processing unit 302 determines the first time range matching the third time range from the at least one first time range, it can be specifically used to: determine the first time subordinate index corresponding to each of the first time ranges, and determine the first time subordinate index corresponding to the third time range; based on the first time subordinate index corresponding to each of the first time ranges and the first time subordinate index corresponding to the third time range, determine the first time range matching the third time range from the at least one first time range, the determined first time range being: the first time range having the same first time subordinate index as the first time subordinate index corresponding to the third time range;

[0122] When the processing unit 302 determines the second time range that matches the third time range from the at least one second time range, it can be specifically used to: determine the second time subordinate indicators corresponding to the respective second time ranges, and determine the second time subordinate indicators corresponding to the third time range; based on the second time subordinate indicators corresponding to the respective second time ranges and the second time subordinate indicators corresponding to the third time range, determine the second time range that matches the third time range from the at least one second time range, the determined second time range being: a second time range whose second time subordinate indicator is the same as the second time subordinate indicator corresponding to the third time range.

[0123] According to another embodiment of the present invention, Figure 3 Each unit in the probability forecasting device shown can be separately or completely combined into one or several other units to form, or one (some) of the units can be further divided into multiple functionally smaller units to form, which can achieve the same operation without affecting the realization of the technical effects of the embodiments of the present invention. The above-mentioned units are divided based on logical functions. In practical applications, the function of one unit can also be realized by multiple units, or the functions of multiple units can be realized by one unit. In other embodiments of the present invention, any probability forecasting device can also include other units. In practical applications, these functions can also be implemented with the assistance of other units, and can be implemented by the collaboration of multiple units.

[0124] According to another embodiment of the present invention, the program can be executed by running a program on a general electronic device such as a computer including a central processing unit (CPU), a random access memory medium (RAM), a read-only memory medium (ROM), and other processing elements and storage elements. Figure 1 or Figure 2 A computer program (including program code) for each step involved in the corresponding method shown in Figure 3 The probability forecasting device shown in the embodiment of the present invention is used to implement the probability forecasting method of the embodiment of the present invention. The computer program can be recorded on a computer storage medium, for example, and loaded into the above-mentioned electronic device through the computer storage medium and run therein.

[0125] Based on the description of the above method embodiment and device embodiment, the exemplary embodiment of the present invention further provides an electronic device, including: at least one processor; and a memory connected to the at least one processor in communication. The memory stores a computer program that can be executed by the at least one processor, and the computer program is used to enable the electronic device to perform the method according to the embodiment of the present invention when executed by the at least one processor.

[0126] Exemplary embodiments of the present invention also provide a non-transitory computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor of a computer, is used to cause the computer to perform a method according to an embodiment of the present invention.

[0127] An exemplary embodiment of the present invention further provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor of a computer, the computer is used to enable the computer to perform a method according to an embodiment of the present invention.

[0128] refer to Figure 4 , a block diagram of an electronic device 400 that can be used as a server or client of the present invention will now be described, which is an example of a hardware device that can be applied to various aspects of the present invention. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0129] like Figure 4 As shown, the electronic device 400 includes a computing unit 401, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 402 or a computer program loaded from a storage unit 408 into a random access memory (RAM) 403. In the RAM 403, various programs and data required for the operation of the electronic device 400 can also be stored. The computing unit 401, the ROM 402, and the RAM 403 are connected to each other via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0130] A plurality of components in the electronic device 400 are connected to the I / O interface 405, including: an input unit 406, an output unit 407, a storage unit 408, and a communication unit 409. The input unit 406 may be any type of device capable of inputting information to the electronic device 400, and the input unit 406 may receive input digital or character information, and generate key signal inputs related to user settings and / or function control of the electronic device. The output unit 407 may be any type of device capable of presenting information, and may include, but is not limited to, a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 408 may include, but is not limited to, a disk, an optical disk. The communication unit 409 allows the electronic device 400 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks, and may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication transceiver, and / or a chipset, such as a Bluetooth™ device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0131] The computing unit 401 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 401 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 401 performs the various methods and processes described above. For example, in some embodiments, the probability forecasting method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 408. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 400 via the ROM 402 and / or the communication unit 409. In some embodiments, the computing unit 401 may be configured to perform the probability forecasting method in any other appropriate manner (e.g., by means of firmware).

[0132] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0133] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0134] As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and / or device (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0135] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0136] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0137] Furthermore, it should be understood that what is disclosed above is only a preferred embodiment of the present invention, and certainly cannot be used to limit the scope of rights of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.

Claims

1. A probabilistic forecasting method, characterized in that: include: Obtain an observation data set and a forecast data set corresponding to each forecast scheme in N forecast schemes, wherein the observation data set includes: an observation event data set of a target area in each first time range in at least one first time range, and a forecast data set includes: a forecast event data set of the target area in each first time range under a corresponding forecast scheme; wherein one forecast event data is obtained by forecasting an event using the forecast scheme corresponding to the forecast data set, and N is a positive integer; one event support is any one of the following: an air quality level event, a primary pollutant event, a precipitation level event, and a wind force level event; Based on the observation data set and the forecast data set corresponding to each forecast scheme, determine the forecast condition probability set of each forecast scheme in each second time range in at least one second time range, wherein the forecast condition probability set of a forecast scheme in a second time range is: the forecast condition probability set obtained by the corresponding forecast scheme for forecasting events in the target area in the corresponding second time range; Based on the observation data set, calculating the observation occurrence probability set within each first time range; and based on the forecast data set corresponding to each forecast scheme, calculating the forecast probability set of each forecast scheme within each first time range; Determining the occurrence probability of each target event in at least one target event under the event forecast result set based on the observation occurrence probability sets within the respective first time ranges, the forecast probability sets of the respective forecast schemes within the respective first time ranges, and the forecast condition probability sets of the respective forecast schemes within the respective second time ranges, comprising: For any target event among the at least one target event, determine a first time range that matches the third time range from the at least one first time range, and select a target observation occurrence probability of the third time range under the any target event from the determined observation occurrence probability set within the first time range; Selecting respectively from the forecast probability sets of the respective forecast schemes within the determined first time range the forecast probabilities of the respective event forecast results within the third time range in the event forecast result set, the event forecast result set including the event forecast results obtained by the respective forecast schemes for forecasting the event for the target area within the third time range; Determine a second time range matching the third time range from the at least one second time range, and select the probability of occurrence of each event forecast result under the condition that any target event has occurred from the forecast condition probability set of each forecast scheme within the determined second time range; The method comprises: calculating the probability of occurrence of any target event in the event forecast result set based on the target observation occurrence probability, the forecast probability of the third time range under the respective event forecast results, and the occurrence probability of the respective event forecast results under the condition that any target event has occurred, including: traversing each event forecast result in the event forecast result set in turn, and taking the currently traversed event forecast result as the current event forecast result; when the current event forecast result is the first event forecast result in the event forecast result set, calculating the probability of occurrence of any target event in the result set corresponding to the current event forecast result based on the target observation occurrence probability, the forecast probability of the third time range under the current event forecast result, and the occurrence probability of the current event forecast result under the condition that any target event has occurred; when the current event forecast result is not the first event forecast result, calculating the probability of occurrence of any target event in the result set corresponding to the current event forecast result. When calculating the prediction result, the occurrence probability of any target event in the result set corresponding to the current event forecast result is calculated based on the occurrence probability of any target event in the result set corresponding to the previous event forecast result of the current event forecast result, the forecast probability of the third time range in the current event forecast result, and the occurrence probability of the current event forecast result under the condition that any target event has occurred. After traversing each event forecast result in the event forecast result set, the occurrence probability of any target event in the event forecast result set is obtained. The occurrence probability of any target event in the event forecast result set is the occurrence probability of any target event in the result set corresponding to the Nth event forecast result in the event forecast result set. The result set corresponding to one event forecast result includes each event forecast result from the first event forecast result to the corresponding event forecast result.

2. The method according to claim 1, characterized in that The determining, based on the observation data set and the forecast data set corresponding to each forecast scheme, a forecast condition probability set of each forecast scheme in each second time range in at least one second time range, comprises: For any second time range of the at least one second time range, determine P first time ranges corresponding to the any second time range, and determine an observation event data set within the P first time ranges from the observation data set, where P is a positive integer; Traversing each forecast scheme in the N forecast schemes, and taking the currently traversed forecast scheme as the current forecast scheme, and determining the P forecast event data sets within the first time range from the forecast data sets corresponding to the current forecast scheme; Using the P observed event data sets and forecast event data sets within the first time range, determine the forecast condition probability set of the current forecast scheme within any second time range; wherein a forecast condition probability set includes a forecast scheme's conditional occurrence probability set under the condition that each observed event in a plurality of observed events within a second time range has occurred, and a conditional occurrence probability set includes the occurrence probability of each forecast event in a plurality of forecast events under the condition that one observed event has occurred; After traversing each forecasting scheme in the N forecasting schemes, a forecasting condition probability set of each forecasting scheme within any second time range is obtained.

3. The method according to claim 2, characterized in that The using the P observed event data sets and the forecast event data sets within the first time range to determine a forecast condition probability set of the current forecast scheme within any second time range includes: Traversing each observation event in the plurality of observation events, and taking the currently traversed observation event as the current observation event, and using the observation event data sets within the P first time ranges to determine the number of occurrences of the current observation event; Using the P forecast event data sets within the first time range, respectively determine the forecast times of each forecast event under the condition that the current observed event has occurred; Calculating the probability of occurrence of each forecast event under the condition that the current observed event has occurred based on the number of occurrences of the current observed event and the number of forecasts of each forecast event under the condition that the current observed event has occurred, so as to obtain a set of conditional occurrence probabilities of the current forecast scheme under the condition that the current observed event has occurred within any second time range; After traversing each of the multiple observation events, a forecast condition probability set of the current forecast scheme within any second time range is obtained.

4. The method according to claim 3, characterized in that The calculating the probability of occurrence of each forecast event under the condition that the current observed event has occurred based on the number of occurrences of the current observed event and the number of forecasts of each forecast event under the condition that the current observed event has occurred, comprises: Determine the conditional smoothing value; The occurrence probability of each forecast event under the condition that the current observed event has occurred is calculated based on the number of occurrences of the current observed event, the conditional smoothing value and the forecast number of each forecast event under the condition that the current observed event has occurred.

5. The method according to any one of claims 1 to 4, characterized in that: An observation event data set is used to indicate the number of occurrences of each observation event among a plurality of observation events within a corresponding first time range, and a forecast event data set is used to indicate the number of forecasts of each forecast event among a plurality of forecast events within a corresponding first time range for a corresponding forecast scheme; The step of calculating the observation occurrence probability sets within each first time range based on the observation data set includes: Calculating observation occurrence probability sets within each first time range based on the number of occurrences indicated by the observation event data sets within each first time range, wherein one observation occurrence probability set includes the observation occurrence probability of each observation event within the corresponding first time range; The calculating, based on the forecast data sets corresponding to the forecast schemes, forecast probability sets of the forecast schemes within the first time ranges respectively, includes: For any forecasting scheme among the N forecasting schemes, a forecast probability set of any forecasting scheme within each first time range is calculated based on the number of forecasts indicated by the forecast event data set of any forecasting scheme within each first time range, and one forecast probability set includes the forecast probability of each forecast event of the corresponding forecasting scheme within the corresponding first time range.

6. The method according to claim 5, characterized in that The calculating the observation occurrence probability sets within each first time range based on the occurrence times indicated by the observation event data sets within each first time range respectively includes: For any first time range of the at least one first time range, determining an observation smoothing value, and calculating an observation occurrence probability set within the any first time range based on the observation smoothing value and the number of occurrences indicated by the observation event data set within the any first time range; The calculating the forecast probability set of any forecast scheme within each first time range based on the forecast times indicated by the forecast event data set of any forecast scheme within each first time range respectively includes: A forecast smoothing value is determined, and based on the forecast smoothing value and the forecast times indicated by the forecast event data set of any forecast scheme within the any first time range, a forecast probability set of any forecast scheme within the any first time range is calculated.

7. The method according to any one of claims 1 to 4, characterized in that: The step of determining, from the at least one first time range, a first time range matching the third time range comprises: Determining the first time-dependent index corresponding to each of the first time ranges, and determining the first time-dependent index corresponding to the third time range; Based on the first time subordinate indicators corresponding to the respective first time ranges and the first time subordinate indicator corresponding to the third time range, a first time range matching the third time range is determined from the at least one first time range, wherein the determined first time range is a first time range having the same first time subordinate indicator as the first time subordinate indicator corresponding to the third time range; The step of determining, from the at least one second time range, a second time range matching the third time range comprises: Determining the second time-dependent index corresponding to each of the second time ranges, and determining the second time-dependent index corresponding to the third time range; Based on the second time subordinate indicators corresponding to the respective second time ranges and the second time subordinate indicators corresponding to the third time range, a second time range matching the third time range is determined from the at least one second time range, and the determined second time range is: a second time range whose second time subordinate indicator is the same as the second time subordinate indicator corresponding to the third time range.

8. A probability forecasting device, characterized in that: The device comprises: An acquisition unit is used to acquire an observation data set and a forecast data set corresponding to each forecast scheme in N forecast schemes, wherein the observation data set includes: an observation event data set of a target area in each first time range in at least one first time range, and a forecast data set includes: a forecast event data set of the target area in each first time range under a corresponding forecast scheme; wherein one forecast event data is obtained by forecasting an event using the forecast scheme corresponding to the forecast data set, and N is a positive integer; one event support is any one of the following: an air quality level event, a primary pollutant event, a precipitation level event, and a wind force level event; A processing unit, configured to determine, based on the observation data set and the forecast data set corresponding to each forecast scheme, a forecast condition probability set of each forecast scheme in each second time range in at least one second time range, wherein the forecast condition probability set of a forecast scheme in a second time range is: a forecast condition probability set obtained by the corresponding forecast scheme performing event forecasting on the target area in the corresponding second time range; The processing unit is further configured to calculate the observation occurrence probability set within each first time range based on the observation data set; and calculate the forecast probability set of each forecast scheme within each first time range based on the forecast data set corresponding to each forecast scheme; The processing unit is further used to determine the occurrence probability of each target event in at least one target event under the event forecast result set based on the observation occurrence probability set within each first time range, the forecast probability set of each forecast scheme within each first time range, and the forecast condition probability set of each forecast scheme within each second time range, including: For any target event among the at least one target event, determine a first time range that matches the third time range from the at least one first time range, and select a target observation occurrence probability of the third time range under the any target event from the determined observation occurrence probability set within the first time range; Selecting respectively from the forecast probability sets of the respective forecast schemes within the determined first time range the forecast probabilities of the respective event forecast results within the third time range in the event forecast result set, the event forecast result set including the event forecast results obtained by the respective forecast schemes for forecasting the event for the target area within the third time range; Determine a second time range matching the third time range from the at least one second time range, and select the probability of occurrence of each event forecast result under the condition that any target event has occurred from the forecast condition probability set of each forecast scheme within the determined second time range; The method comprises: calculating the probability of occurrence of any target event in the event forecast result set based on the target observation occurrence probability, the forecast probability of the third time range under the respective event forecast results, and the occurrence probability of the respective event forecast results under the condition that any target event has occurred, including: traversing each event forecast result in the event forecast result set in turn, and taking the currently traversed event forecast result as the current event forecast result; when the current event forecast result is the first event forecast result in the event forecast result set, calculating the probability of occurrence of any target event in the result set corresponding to the current event forecast result based on the target observation occurrence probability, the forecast probability of the third time range under the current event forecast result, and the occurrence probability of the current event forecast result under the condition that any target event has occurred; when the current event forecast result is not the first event forecast result, calculating the probability of occurrence of any target event in the result set corresponding to the current event forecast result. When calculating the prediction result, the occurrence probability of any target event in the result set corresponding to the current event forecast result is calculated based on the occurrence probability of any target event in the result set corresponding to the previous event forecast result of the current event forecast result, the forecast probability of the third time range in the current event forecast result, and the occurrence probability of the current event forecast result under the condition that any target event has occurred. After traversing each event forecast result in the event forecast result set, the occurrence probability of any target event in the event forecast result set is obtained. The occurrence probability of any target event in the event forecast result set is the occurrence probability of any target event in the result set corresponding to the Nth event forecast result in the event forecast result set. The result set corresponding to one event forecast result includes each event forecast result from the first event forecast result to the corresponding event forecast result.

9. An electronic device, characterized in that: include: processor; as well as Memory for storing programs, The program includes instructions, which, when executed by the processor, cause the processor to perform the method according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to make a computer execute the method according to any one of claims 1-7.