Probability calculation device, probability calculation method, and probability calculation program

The probability calculation device aggregates data from similar regions using clustering and vector analysis to address the challenge of limited data in subdivided regions, enhancing accuracy and reducing errors in event occurrence predictions.

JP7578195B2Active Publication Date: 2024-11-06NIPPON TELEGRAPH & TELEPHONE CORP
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
JP2023529331
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2021-06-23
Publication Date
2024-11-06
Estimated Expiration
2041-06-23

AI Technical Summary

Technical Problem

Existing methods struggle to accurately calculate the probability of event occurrence in subdivided regions with limited data, leading to increased errors and inaccuracies.

Method used

A probability calculation device and method that aggregates event occurrence data from similar regions, using clustering and vector calculations to suppress errors and enhance accuracy, even with limited data.

Benefits of technology

Enables precise calculation of event occurrence probabilities in subdivided regions by reducing the impact of limited data through clustering and vector analysis, thereby improving prediction accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007578195000001
    Figure 0007578195000001
  • Figure 0007578195000002
    Figure 0007578195000002
  • Figure 0007578195000003
    Figure 0007578195000003
Patent Text Reader

Abstract

A probability calculation device according to the present invention calculates, for each target and on the basis of a basic occurrence probability and a conditional occurrence probability, a first fluctuation vector as a vector expressing a fluctuation corresponding to a specified explanatory variable, said calculation being performed in a first fluctuation calculation unit. The targets are classified into clusters as a result of a clustering unit clustering respective fluctuation vectors of the targets. For each cluster, a second fluctuation calculation unit aggregates event occurrence data which corresponds to the targets belonging to said cluster, and calculates a second fluctuation vector as a vector expressing a fluctuation corresponding to a specified explanatory variable for said cluster. For each vector, an integrated calculation unit uses the second fluctuation vector of the cluster to which the target belongs to calculate the event occurrence probability for the target in a case in which the specific explanatory variable serves as a condition.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The disclosed technology relates to a probability calculation device, a probability calculation method, and a probability calculation program. [Background technology]

[0002] There are cases where it is desired to predict the probability of some event occurrence, such as the occurrence of an accident or the occurrence of an injured person, for each subdivided region based on past event occurrence data, taking into account the fact that it varies depending on several conditions, such as the day of the week or the time of day. In other words, there are cases where it is desired to obtain a conditional event occurrence probability. For example, there are cases where it is desired to pre-deploy emergency vehicles, such as police cars or ambulances, to appropriate locations based on the obtained event occurrence probability. Non-Patent Document 1 describes a method of optimizing the deployment of ambulances by performing real-time emergency demand prediction. [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] "Verifying the effectiveness of an optimal ambulance operation system using emergency big data" URL: https: / / www.ntt.co.jp / news2018 / 1811 / 181126a.html Summary of the Invention [Problem to be solved by the invention]

[0004] For the above-mentioned purposes, it is necessary to divide the target area into a number of small areas and calculate the event occurrence probability for each small area, rather than calculating the event occurrence probability for the entire area. However, dividing the entire area into small areas reduces the event occurrence probability for each small area.

[0005] The disclosed technology has been made in consideration of the above-mentioned points, and aims to provide a probability calculation device, a probability calculation method, and a probability calculation program that can appropriately calculate the probability of an event occurring in a target while suppressing errors even when there is a small amount of observed data about the target. [Means for solving the problem]

[0006] A first aspect of the present disclosure is a probability calculation device including: a first variation calculation unit that calculates a first variation vector for each of a plurality of targets as a vector representing a variation according to a specific explanatory variable based on a basic occurrence probability, which is an occurrence probability obtained from past event occurrence data for each of the targets, and a conditional occurrence probability, which is an occurrence probability conditioned on a specific explanatory variable of the event occurrence data; a clustering unit that classifies the targets into clusters by clustering the variation vectors of each of the targets; a second variation calculation unit that aggregates, for each cluster, the event occurrence data associated with the targets belonging to the cluster, and calculates a second variation vector as a vector representing a variation according to the specific explanatory variable of the cluster; and an integration calculation unit that calculates, for each of the targets, the event occurrence probability of the target using the second variation vector of the cluster to which the target belongs, when the specific explanatory variable is used as a condition.

[0007] A second aspect of the present disclosure is a probability calculation method that causes a computer to execute a process of calculating, for each of a plurality of targets, a first variation vector as a vector representing variation according to a specific explanatory variable based on a basic occurrence probability, which is an occurrence probability obtained from past event occurrence data for each of the targets, and a conditional occurrence probability, which is an occurrence probability conditioned on a specific explanatory variable of the event occurrence data, classifying the targets into clusters by clustering each of the variation vectors of the targets, aggregating, for each cluster, the event occurrence data associated with the targets belonging to the cluster, calculating a second variation vector as a vector representing variation according to the specific explanatory variable of the cluster, and calculating, for each target, the event occurrence probability of the target when the specific explanatory variable is used as a condition, using the second variation vector of the cluster to which the target belongs.

[0008] A third aspect of the present disclosure is a probability calculation program that causes a computer to execute the following processes: calculate a first variation vector for each of a plurality of targets as a vector representing variation according to a specific explanatory variable based on a basic occurrence probability, which is an occurrence probability obtained from past event occurrence data for each of the targets, and a conditional occurrence probability, which is an occurrence probability conditioned on a specific explanatory variable of the event occurrence data; classify the targets into clusters by clustering the variation vectors of each of the targets; aggregate, for each cluster, the event occurrence data associated with the targets belonging to the cluster; calculate a second variation vector as a vector representing variation according to the specific explanatory variable of the cluster; and calculate, for each target, the event occurrence probability of the target using the second variation vector of the cluster to which the target belongs when the specific explanatory variable is used as a condition. Effect of the Invention

[0009] According to the disclosed technology, even when there is a small amount of observed data about a target, it is possible to appropriately calculate the probability of an event occurring for the target while suppressing errors. [Brief description of the drawings]

[0010] [Figure 1] FIG. 13 is a diagram showing an example of a data string generated on the assumption that regions are subdivided into small regions. [Diagram 2] 1 is a block diagram showing a hardware configuration of a probability calculation device according to an embodiment of the present disclosure. [Diagram 3] 1 is a block diagram illustrating a functional configuration of a probability calculation device according to an embodiment of the present disclosure. [Figure 4] FIG. 13 is a diagram showing an example of a graph showing the results of determining first variation vectors for a plurality of small regions. [Diagram 5] FIG. 13 is a diagram showing the results of clustering the first change vectors for each small region into two groups using k-means, and determining the second change vectors for each cluster. [Figure 6] 11 is a flowchart showing a flow of a probability calculation process performed by the probability calculation device according to the embodiment of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, an example of an embodiment of the disclosed technology will be described with reference to the drawings. In each drawing, the same or equivalent components and parts are given the same reference numerals. In addition, the dimensional ratios of the drawings are exaggerated for the convenience of explanation and may differ from the actual ratios.

[0012] First, an overview of the present disclosure will be described. As described above, when an entire region is divided into smaller regions, the probability of an event occurring in each small region becomes smaller. For convenience of explanation, the probability of an event that is ultimately desired to be obtained in this embodiment will be referred to as the event occurrence probability, and the probability of occurrence of other events will be simply referred to as the occurrence probability.

[0013] When the occurrence probability is small, more event occurrence data (observation data) is required, but cases where the amount of data is insufficient are assumed. For example, if the true occurrence probability of injuries and illnesses in the entire region is 0.1, a computer generates a binary random number each time according to this occurrence probability, and if the value is 0, it is assumed that the event did not occur, and if the value is 1, it is assumed that the event occurred. In order to stably obtain the original occurrence probability from the observation result (a data string consisting of 0 and 1), the random number generation algorithm usually requires approximately 100 observations. Here, it is assumed that the true occurrence probability in a certain small region is 0.01 as a result of dividing the entire region into small regions. In that case, approximately 1,000 occurrences of event occurrence data are required. Figure 1 shows an example of a data string generated under the assumption that the region is divided into small regions. If 1000 observations are made, the occurrence probability is 0 / 10 = 0 if only A1 in the data string is observed, and 1 / 10 = 0.1 if only A2 is observed, meaning that 1 will appear about 10 times in the 1000 data strings. In actual observations, it is expected that it will be difficult to obtain that much event occurrence data, so it becomes necessary to calculate the occurrence probability of a small region from event occurrence data with a small number of observations. However, when the occurrence probability of a small region is calculated from event occurrence data with a small number of observations, in many cases the occurrence probability will be 0, or there is a possibility that a probability greater than the true occurrence probability will be erroneously calculated.

[0014] Furthermore, if one tries to find a conditional probability of an event occurring, it is expected that the shortage of event occurrence data will become even more pronounced. For example, if one wishes to find the variation in the probability of injury or illness occurring depending on the day of the week, the data string of event occurrence data must be divided into seven parts. In particular, if one assumes that the influence of conditions such as the day of the week only varies the base occurrence probability by about 20-30%, this would mean that the occurrence probability of 0.01 would vary to 0.012 or 0.013, etc., making the shortage of data even more pronounced.

[0015] In the above example, some of the subdivided small areas may have similar urban functions, such as small areas known as residential areas or entertainment districts. For example, in entertainment districts, there is a possibility that the number of people suffering from acute alcohol poisoning increases on certain days of the week and at certain times of the day, and the probability of people suffering from illness increases in all entertainment districts. In residential areas, the probability of people suffering from illness increases on days and times when many people are at home.

[0016] In this way, it can be assumed that small regions with similar urban functions, even if they are not adjacent, have similar effects on the probability of occurrence due to the conditions of specific explanatory variables. Therefore, by aggregating past event occurrence data for small regions, it is possible to first increase the scale of occurrence probability, and then calculate the fluctuation due to specific explanatory variables, and then use this when calculating the fluctuation of each small region. This should make it easier to obtain probabilities close to the true occurrence probability for each of several small regions, even from a relatively small amount of event occurrence data.

[0017] As a method for aggregating small regions with similar urban functions, a vector (first change vector) that represents the fluctuations of multiple regions is calculated from the conditional occurrence probability for each small region, using specific explanatory variables as a condition, and then clustered. By clustering regions based on the vector that represents the fluctuations of each small region, it is possible to negate some observational errors, even if the individual fluctuations in the subdivided small regions cannot be calculated accurately due to the small amount of observational data.

[0018] The reason for clustering using vectors that represent fluctuations is that the absolute occurrence (probability) differs for each small area. If clustering is performed using the above method, for example, an entertainment district in a city center with a particularly high population density and an entertainment district in a small area adjacent to the city center can be classified into the same cluster.

[0019] In the above explanation, in order to easily explain the application of the assumed small areas, the typical image of a city such as an entertainment district or residential area is applied to the small areas, but in reality, many small areas have houses and restaurants scattered throughout and are likely to have complex attributes. Therefore, since it is not easy to determine whether one small area shows the same fluctuations as another, it is important to calculate vectors that represent the fluctuations for each small area and then perform clustering.

[0020] In addition to the above, it is also possible to assume a case where temperature is used as a specific explanatory variable for each of a number of small regions with differences in elevation above sea level. In this case, if small regions with similar elevations can be clustered together, it becomes easier to predict fluctuations in the number of occurrences of heat stroke and the like from limited event occurrence data. Each of the above-mentioned multiple small regions is an example of each of a number of targets of the technology disclosed herein.

[0021] In this embodiment, the case where the event occurrence probability is calculated for a small area will be described as an example, but the scope of application of this embodiment is not limited to small areas in which event occurrence data can be collected in real space. As long as event occurrence data can be collected, the method of this embodiment can be applied regardless of the target, such as the Internet space or a network environment.

[0022] In consideration of the above, the above idea is realized by configuring a probability calculation device according to an embodiment of the present disclosure as follows.

[0023] The configuration of this embodiment will be described below.

[0024] FIG. 2 is a block diagram showing a hardware configuration of the probability calculation device 100 according to the embodiment of the present disclosure.

[0025] 2, the probability calculation device 100 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, a storage 14, an input unit 15, a display unit 16, and a communication interface (I / F) 17. Each component is connected to each other via a bus 19 so as to be able to communicate with each other.

[0026] The CPU 11 is a central processing unit, and executes various programs and controls each part. That is, the CPU 11 reads a program from the ROM 12 or the storage 14, and executes the program using the RAM 13 as a working area. The CPU 11 controls each of the above components and performs various arithmetic processing according to the program stored in the ROM 12 or the storage 14. In this embodiment, the ROM 12 or the storage 14 stores an occurrence probability calculation program.

[0027] The ROM 12 stores various programs and various data. The RAM 13 temporarily stores programs or data as a working area. The storage 14 is configured with a storage device such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive), and stores various programs including an operating system, and various data.

[0028] The input unit 15 includes a pointing device such as a mouse, and a keyboard, and is used to perform various input operations.

[0029] The display unit 16 is, for example, a liquid crystal display, and displays various information. The display unit 16 may function as the input unit 15 by adopting a touch panel system.

[0030] The communication interface 17 is an interface for communicating with other devices such as terminals. For this communication, for example, a wired communication standard such as Ethernet (registered trademark) or FDDI, or a wireless communication standard such as 4G, 5G, or Wi-Fi (registered trademark) is used.

[0031] Next, a description will be given of each functional configuration of the probability calculation device 100. Fig. 3 is a block diagram showing the functional configuration of the probability calculation device 100 according to an embodiment of the present disclosure. Each functional configuration is realized by the CPU 11 reading out an occurrence probability calculation program stored in the ROM 12 or storage 14, expanding it in the RAM 13, and executing it.

[0032] As shown in FIG. 3, the probability calculation device 100 includes an event memory unit 102, a basic probability calculation unit 110, a conditional probability calculation unit 112, a first variation calculation unit 114, a clustering unit 116, a second variation calculation unit 118, and an integrated calculation unit 120.

[0033] The probability calculation device 100 receives event occurrence data as an input, and stores it in the event storage unit 102. The event occurrence data includes the date and time of an event occurring for a certain period of time in the past for each of a plurality of targets, the day of the week, the small area in which the event occurred, and the like.

[0034] For simplicity, in the following explanation, it is assumed that the probability of an event occurring is constant from the past to the future when the time of occurrence, day of the week, and small area are the same. In other words, it is assumed that if enough event occurrence data under the same conditions is accumulated, the probability of an event occurring under the same conditions in the future can be calculated. However, it is assumed that there is not enough event occurrence data accumulated for the scale of the event occurrence probability that is desired.

[0035] The basic probability calculation unit 110 calculates the occurrence probability (basic occurrence probability) for each small area and for each time period based on the event occurrence data stored in the event memory unit 102. The basic occurrence probability can be calculated by counting the number of events in a specific small area and in a specific time period during a certain past period of the event occurrence data, and dividing the number by the number of days in that period. For example, if the period is 350 days and a total of 35 events occurred in a certain small area between 10:00 and 350, the basic occurrence probability is 0.1.

[0036] The conditional probability calculation unit 112 calculates the occurrence probability (conditional occurrence probability) for each small area by day of the week and by time period based on the event occurrence data stored in the event storage unit 102. The conditional occurrence probability can be calculated by counting events in a specific small area and in a specific time period in a past fixed period of the event occurrence data, and dividing the result by the number of days in the fixed period. For example, if a period is 350 days and a total of 35 events occurred in a certain small area between 10:00 and 350 days, the conditional occurrence probability is 0.1. Using the day of the week as a condition is an example of using a specific explanatory variable of the technology disclosed herein as a condition. Note that the probability calculation device 100 may receive the basic occurrence probability and the conditional occurrence probability calculated in advance by another device, etc., and in that case, the basic probability calculation unit 110 and the conditional probability calculation unit 112 may not be required in the configuration.

[0037] The first variation calculation unit 114 receives the input of the basic occurrence probability from the basic probability calculation unit 110 and the conditional occurrence probability from the conditional probability calculation unit 112. The first variation calculation unit 114 calculates a first variation vector for each small area based on the basic occurrence probability and the conditional occurrence probability. The first variation vector is a vector that represents a variation according to a specific explanatory variable, here, the day of the week. The first variation vector here is obtained by calculating how many times the conditional occurrence probability for each day of the week is the basic occurrence probability ignoring the day of the week for each small area. In the above example, the variation on Sundays between 10:00 and 20:00 is 0.06 / 0.1=0.6 times, and the variation is obtained for all combinations of all time periods and all days of the week, and this is set as the first variation vector. FIG. 4 is a diagram showing an example of a graph of the result of obtaining the first variation vector for a plurality of small areas. In the example of FIG. 4, the variation amount for each day of the week and time period for each of small areas A, B, C, and D is obtained as the first variation vector.

[0038] The clustering unit 116 receives an input of the variation vector of each small region from the first variation calculation unit 114. The clustering unit 116 classifies the small regions into clusters by clustering the first variation vector of each small region. A representative clustering method such as k-means may be used. For example, the number of clusters may be a parameter that is manually adjusted, and several numbers of clusters may be tried and finally adjusted so that the future prediction results are most suitable.

[0039] The second variation calculation unit 118 receives the result of classification of each small area into each cluster from the clustering unit 116 as a clustering result. The second variation calculation unit 118 aggregates the event occurrence data associated with the small areas belonging to each cluster for each cluster, and calculates a second variation vector. Specifically, when calculating the second variation vector, the second variation calculation unit 118 acquires and aggregates the event occurrence data of the small areas included in each cluster from the event storage unit 102 for each cluster. The second variation calculation unit 118 uses the aggregated event occurrence data for the cluster to obtain a basic occurrence probability of the cluster in a time period, and obtains a conditional occurrence probability of the cluster using the occurrence probability by day of the week and by time period as a condition. Then, the second variation vector is calculated based on the basic occurrence probability of the cluster and the conditional occurrence probability of the cluster. This makes it possible to calculate the second variation vector as a vector representing a variation according to a specific explanatory variable of the cluster.

[0040] FIG. 5 is a diagram showing the result of clustering the first fluctuation vector for each small region into two by k-means and obtaining the second fluctuation vector for each cluster. When graphed, the graph is basically smoother than the original first fluctuation vector for each small region. In addition, since the second fluctuation vector is recalculated after aggregating the event occurrence data of the clustered small regions rather than directly obtaining the average vector from the fluctuation vector for each small region, the influence of the original small region with a large basic occurrence probability is stronger. This effectively weakens the influence of the small region with a large observation error. Although a similar result may be obtained by widening the time span in the original small region, the method of the present disclosure is appropriate when there is a place where the fluctuation of the true occurrence probability increases and decreases every hour.

[0041] The integration calculation unit 120 receives the input of the second variation vector for each cluster from the second variation calculation unit 118. The integration calculation unit 120 calculates the event occurrence probability for each small area based on the basic occurrence probability for each time period of the small area and the variation value of the second variation vector of the cluster to which the small area belongs. The event occurrence probability is calculated as a conditional occurrence probability for each small area. For example, to calculate the event occurrence probability for each small area on Sunday between 10:00 and 10:00, the basic occurrence probability for each small area is calculated for each small area regardless of the day of the week. After that, the variation value for Sunday between 10:00 and 10:00 is extracted from the second variation vector of the cluster to which the small area belongs, and multiplied to calculate the conditional occurrence probability for Sunday between 10:00 and 10:00 for each small area as the event occurrence probability.

[0042] Next, the operation of the probability calculation device 100 will be described.

[0043] 6 is a flowchart showing the flow of the probability calculation process by the probability calculation device 100 according to the embodiment of the present disclosure. The probability calculation process is performed by the CPU 11 reading out the probability calculation program from the ROM 12 or storage 14, expanding it in the RAM 13, and executing it. The CPU 11 executes the following processes as each part of the probability calculation device 100.

[0044] In step S100, the CPU 11, functioning as the basic probability calculation unit 110, calculates occurrence probabilities (basic occurrence probabilities) for each time period for each small area, based on the event occurrence data stored in the event storage unit .

[0045] In step S102, the CPU 11, functioning as the conditional probability calculation unit 112, calculates occurrence probabilities (conditional occurrence probabilities) for each small area by day of the week and by time period, based on the event occurrence data stored in the event storage unit 102.

[0046] In step S104, the CPU 11, functioning as the first variation calculation unit 114, calculates a first variation vector for each small region based on the basic occurrence probability and the conditional occurrence probability. The first variation vector is a vector that represents a variation according to a specific explanatory variable, here, the day of the week.

[0047] In step S106, CPU 11, functioning as clustering section 116, classifies the small regions into clusters by clustering the first variation vectors of the small regions.

[0048] In step S108, CPU 11, functioning as second variation calculation section 118, aggregates, for each cluster, event occurrence data associated with small areas belonging to that cluster, and calculates a second variation vector.

[0049] In step S110, the CPU 11, as the integrated calculation unit 120, calculates an event occurrence probability for each small area based on the basic occurrence probability for each time period of the small area and the fluctuation value of the second fluctuation vector of the cluster to which the small area belongs.

[0050] As described above, according to the probability calculation device 100 of this embodiment, even if there is little observed data for a subdivided small area, it is possible to appropriately calculate the event occurrence probability for the small area while suppressing errors.

[0051] In the above embodiment, the probability calculation process executed by the CPU by reading the software (program) may be executed by various processors other than the CPU. In this case, examples of the processor include a PLD (Programmable Logic Device) such as an FPGA (Field-Programmable Gate Array) whose circuit configuration can be changed after manufacture, and a dedicated electric circuit such as an ASIC (Application Specific Integrated Circuit) which is a processor having a circuit configuration designed exclusively for executing a specific process. The probability calculation process may be executed by one of these various processors, or may be executed by a combination of two or more processors of the same or different types (for example, a plurality of FPGAs, and a combination of a CPU and an FPGA). The hardware structure of these various processors is, more specifically, an electric circuit that combines circuit elements such as semiconductor elements.

[0052] In the above embodiment, the probability calculation program is stored (installed) in advance in the storage 14, but the present invention is not limited to this. The program may be provided in a form stored in a non-transitory storage medium such as a CD-ROM (Compact Disk Read Only Memory), a DVD-ROM (Digital Versatile Disk Read Only Memory), or a USB (Universal Serial Bus) memory. The program may also be downloaded from an external device via a network.

[0053] The following supplementary notes are further disclosed regarding the above embodiment.

[0054] (Additional note 1) Memory, at least one processor coupled to the memory; Including, The processor, calculating a first variation vector for each of a plurality of targets as a vector representing a variation according to a specific explanatory variable based on a basic occurrence probability, which is an occurrence probability calculated from past event occurrence data for each of the targets, and a conditional occurrence probability, which is an occurrence probability conditioned on a specific explanatory variable of the event occurrence data; classifying the objects into clusters by clustering the variation vectors of each of the objects; aggregating the event occurrence data associated with the objects belonging to each cluster, and calculating a second variation vector as a vector representing a variation according to the specific explanatory variable of each cluster; calculating, for each of the objects, a probability of an event occurring for the object when the specific explanatory variable is set as a condition, using the second variation vector of the cluster to which the object belongs; The probability calculation device is configured as follows.

[0055] (Additional note 2) A non-transitory storage medium storing a program executable by a computer to execute a probability calculation process, calculating a first variation vector for each of a plurality of targets as a vector representing a variation according to a specific explanatory variable based on a basic occurrence probability, which is an occurrence probability calculated from past event occurrence data for each of the targets, and a conditional occurrence probability, which is an occurrence probability conditioned on a specific explanatory variable of the event occurrence data; classifying the objects into clusters by clustering the variation vectors of each of the objects; aggregating the event occurrence data associated with the objects belonging to each cluster, and calculating a second variation vector as a vector representing a variation according to the specific explanatory variable of each cluster; calculating, for each of the objects, a probability of an event occurring for the object when the specific explanatory variable is set as a condition, using the second variation vector of the cluster to which the object belongs; Non-transitory storage medium. [Explanation of symbols]

[0056] 100 Probability Calculator 102 Event Memory Unit 110 Basic Probability Calculation Unit 112 Conditional Probability Calculation Unit 114 First fluctuation calculation unit 116 Clustering Department 118 Second Fluctuation Calculation Unit 120 Integrated Calculation Department

Claims

1. a first variation calculation unit that calculates, for each of a plurality of targets, a first variation vector as a vector representing a variation according to a specific explanatory variable based on a basic occurrence probability, which is an occurrence probability calculated from past event occurrence data for each of the targets, and a conditional occurrence probability, which is an occurrence probability conditioned on a specific explanatory variable of the event occurrence data; a clustering unit that classifies the objects into clusters by clustering the first variation vectors of each of the objects; a second variation calculation unit that aggregates, for each cluster, the event occurrence data associated with the targets belonging to the cluster, and calculates a second variation vector as a vector representing a variation according to the specific explanatory variable of the cluster; an integration calculation unit that calculates, for each of the objects, an event occurrence probability of the object when the specific explanatory variable is set as a condition, using the second variation vector of the cluster to which the object belongs; A probability calculation device comprising:

2. The probability calculation device according to claim 1 , wherein the integrated calculation unit calculates the event occurrence probability for each of the objects based on the basic occurrence probability of the object and a fluctuation value in the second fluctuation vector of the cluster to which the object belongs.

3. 3. The probability calculation device according to claim 1, wherein the second variation calculation unit calculates, for each cluster, the basic occurrence probability of the cluster using the event occurrence data aggregated for the cluster, and also calculates the conditional occurrence probability of the cluster using the specific explanatory variable, and calculates the second variation vector based on the basic occurrence probability of the cluster and the conditional occurrence probability of the cluster.

4. calculating a first variation vector for each of a plurality of targets as a vector representing a variation according to a specific explanatory variable based on a basic occurrence probability, which is an occurrence probability calculated from past event occurrence data for each of the targets, and a conditional occurrence probability, which is an occurrence probability conditioned on a specific explanatory variable of the event occurrence data; classifying the objects into clusters by clustering the first variation vectors for each of the objects; aggregating the event occurrence data associated with the objects belonging to each cluster, and calculating a second variation vector as a vector representing a variation according to the specific explanatory variable of each cluster; calculating, for each of the objects, a probability of an event occurring for the object when the specific explanatory variable is set as a condition, using the second variation vector of the cluster to which the object belongs; A method for calculating the probability of having a computer execute a process.

5. calculating a first variation vector for each of a plurality of targets as a vector representing a variation according to a specific explanatory variable based on a basic occurrence probability, which is an occurrence probability calculated from past event occurrence data for each of the targets, and a conditional occurrence probability, which is an occurrence probability conditioned on a specific explanatory variable of the event occurrence data; classifying the objects into clusters by clustering the first variation vectors for each of the objects; aggregating the event occurrence data associated with the objects belonging to each cluster, and calculating a second variation vector as a vector representing a variation according to the specific explanatory variable of each cluster; calculating, for each of the objects, a probability of an event occurring for the object when the specific explanatory variable is set as a condition, using the second variation vector of the cluster to which the object belongs; A probability calculation program that causes a computer to execute processing.

Citation Information

Patent Citations

  • Emergency work simulation system and method

    JP2005284991A

  • Risk case display program, risk case display method, and risk case display device

    JP2016042369A

  • Estimate of damage prevention with architectural structure renovation

    JP2019108798A