Construction method of risk prediction model, risk prediction method and device, computer equipment, storage medium and computer program product

Abstract

The invention relates to a risk prediction model construction method and device, equipment, a storage medium and a program product, and the method comprises the steps: carrying out the data expansion according to the distribution condition of historical data of a vehicle, obtaining sample data, inputting the sample data into an initial risk prediction model, carrying out the risk prediction, and obtaining a prediction result. And finally, adjusting the initial risk prediction model according to a prediction result to obtain a target risk prediction model. According to the method, data expansion is performed on the historical data of the vehicle, a large amount of sample data can be obtained, then the risk prediction model is optimized based on the large amount of sample data, the problem of inaccurate prediction caused by insufficient sample data amount in a traditional method can be avoided, and the accuracy of the risk prediction model is further improved. Meanwhile, data expansion is performed according to the distribution condition of the historical data of the vehicle, so that the model optimization efficiency can be improved to a certain extent.

Description

Technical Field

[0001] The present application relates to the technical field of big data analysis, and in particular to a method for constructing a risk prediction model, a risk prediction method, an apparatus, a computer device, a storage medium, and a computer program product. Background Art

[0002] With the development of network information, big data analysis has the advantages of high efficiency, accuracy and real-time, and has been applied in many fields, especially in vehicle risk assessment.

[0003] At present, a vehicle risk prediction model can usually be built based on algorithms such as neural networks or machine learning to analyze and predict the correlation between vehicles and traffic accidents, so as to reduce the occurrence of major traffic accidents.

[0004] However, current risk prediction models suffer from inaccuracies. Summary of the invention

[0005] Based on this, it is necessary to provide a method for constructing a risk prediction model, a risk prediction method, an apparatus, a computer device, a storage medium and a computer program product that can improve the prediction accuracy in response to the above-mentioned technical problems.

[0006] In a first aspect, the present application provides a method for constructing a risk prediction model, the method comprising:

[0007] Perform data expansion based on the distribution of historical data of vehicles to obtain sample data;

[0008] Input the sample data into the initial risk prediction model to perform risk prediction and obtain the prediction results;

[0009] The initial risk prediction model is adjusted according to the prediction results to obtain the target risk prediction model.

[0010] The method for constructing a risk prediction model provided in an embodiment of the present application performs data expansion according to the distribution of the historical data of the vehicle to obtain sample data, and then inputs the sample data into the initial risk prediction model to perform risk prediction to obtain a prediction result, and finally adjusts the initial risk prediction model according to the prediction result to obtain a target risk prediction model. In the above method, data expansion is performed on the historical data of the vehicle to obtain a large amount of sample data, and then the risk prediction model is optimized based on the large amount of sample data, which can avoid the problem of inaccurate prediction caused by insufficient sample data in traditional methods, thereby improving the accuracy of the risk prediction model. At the same time, data expansion according to the distribution of the historical data of the vehicle can improve the efficiency of model optimization to a certain extent.

[0011] In one embodiment, data expansion is performed based on the distribution of historical data of vehicles to obtain sample data, including:

[0012] Classify historical data to obtain multiple types of historical data;

[0013] Data expansion is performed based on the distribution of various types of historical data to obtain sample data.

[0014] In one embodiment, data expansion is performed according to the distribution of various types of historical data to obtain sample data, including:

[0015] According to the distribution of various types of historical data, the distribution relationship corresponding to each distribution situation is determined;

[0016] The data randomly generated according to the preset sample quantity and various distribution relationship formulas constitute multiple groups of sample data; each group of sample data includes all types of data.

[0017] The method described in the embodiments of the present application, by fitting the distribution relationship corresponding to each type of historical data and then expanding the data based on the distribution relationship, can not only obtain a large amount of sample data, but also make the generated sample data close to the real data, which can further improve the prediction accuracy of the model.

[0018] In one embodiment, the prediction result includes a predicted accident rate, and the sample data is input into an initial risk prediction model for risk prediction to obtain a prediction result, including:

[0019] Input the sample data into the initial risk prediction model to perform risk prediction, and obtain the risk value corresponding to each group of sample data;

[0020] Accident assessment is performed based on the risk value corresponding to each group of sample data to obtain the predicted accident rate.

[0021] The method described in the embodiment of the present application predicts the accident rate of the vehicle through the risk value, and the initial risk prediction model can be adjusted based on the predicted accident rate, so that the adjusted model can have a higher prediction accuracy.

[0022] In one embodiment, the initial risk prediction model is adjusted according to the prediction result to obtain a target risk prediction model, including:

[0023] Verify the prediction results to obtain verification results;

[0024] If the verification result indicates that the verification is passed, the initial risk prediction model is determined as the target risk prediction model;

[0025] If the verification result indicates that the verification has not passed, the parameters in the initial risk prediction model are adjusted to obtain the target risk prediction model.

[0026] In one embodiment, the prediction result is verified to obtain a verification result, including:

[0027] Get some data from historical data as verification data;

[0028] Conduct accident assessment on the verification data to obtain the actual accident rate;

[0029] The prediction results are verified according to the actual accident rate to obtain the verification results.

[0030] The method described in the embodiments of the present application can obtain a target risk prediction model that can accurately predict by verifying the prediction results and then adjusting the risk function in the initial risk prediction model based on the verification results.

[0031] In a second aspect, the present application provides a risk prediction method, the method comprising:

[0032] Obtain accident data of multiple accident vehicles;

[0033] Input each accident data into the target risk prediction model for risk prediction, and obtain the risk value corresponding to each accident data;

[0034] Analyze the risk value and the corresponding accident data corresponding to each accident data to determine the associated factors of the vehicle accident.

[0035] The method described in the embodiment of the present application, since the target risk prediction model can accurately predict the risk value corresponding to the accident data of the accident vehicle, the influencing factors of the vehicle accident can be accurately analyzed based on the risk value corresponding to each accident data.

[0036] In a third aspect, the present application also provides a device for constructing a risk prediction model, the device comprising:

[0037] An expansion module is used to expand data according to the distribution of historical data of vehicles to obtain sample data;

[0038] The first prediction module is used to input sample data into the initial risk prediction model to perform risk prediction and obtain prediction results;

[0039] The adjustment module is used to adjust the initial risk prediction model according to the prediction results to obtain the target risk prediction model.

[0040] In a fourth aspect, the present application further provides a risk prediction device, the device comprising:

[0041] An acquisition module, used to acquire accident data of multiple accident vehicles;

[0042] The second prediction module is used to input each accident data into the target risk prediction model for risk prediction to obtain the risk value corresponding to each accident data;

[0043] The analysis module is used to analyze the risk values ​​corresponding to each accident data and the corresponding accident data to determine the associated factors of the vehicle accident.

[0044] In a fifth aspect, the present application further provides a computer device, the computer device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the following steps when executing the computer program:

[0045] Perform data expansion based on the distribution of historical data of vehicles to obtain sample data;

[0046] Input the sample data into the initial risk prediction model to perform risk prediction and obtain the prediction results;

[0047] The initial risk prediction model is adjusted according to the prediction results to obtain the target risk prediction model.

[0048] In a sixth aspect, the present application further provides a computer device, the computer device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the following steps when executing the computer program:

[0049] Obtain accident data of multiple accident vehicles;

[0050] Input each accident data into the target risk prediction model for risk prediction, and obtain the risk value corresponding to each accident data;

[0051] Analyze the risk value and the corresponding accident data corresponding to each accident data to determine the associated factors of the vehicle accident.

[0052] In a seventh aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0053] Perform data expansion based on the distribution of historical data of vehicles to obtain sample data;

[0054] Input the sample data into the initial risk prediction model to perform risk prediction and obtain the prediction results;

[0055] The initial risk prediction model is adjusted according to the prediction results to obtain the target risk prediction model.

[0056] In an eighth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0057] Obtain accident data of multiple accident vehicles;

[0058] Input each accident data into the target risk prediction model for risk prediction, and obtain the risk value corresponding to each accident data;

[0059] Analyze the risk value and the corresponding accident data corresponding to each accident data to determine the associated factors of the vehicle accident.

[0060] In a ninth aspect, the present application further provides a computer program product, the computer program product comprising a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0061] Perform data expansion based on the distribution of historical data of vehicles to obtain sample data;

[0062] Input the sample data into the initial risk prediction model to perform risk prediction and obtain the prediction results;

[0063] The initial risk prediction model is adjusted according to the prediction results to obtain the target risk prediction model.

[0064] In a tenth aspect, the present application further provides a computer program product, the computer program product comprising a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0065] Obtain accident data of multiple accident vehicles;

[0066] Input each accident data into the target risk prediction model for risk prediction, and obtain the risk value corresponding to each accident data;

[0067] Analyze the risk value and the corresponding accident data corresponding to each accident data to determine the associated factors of the vehicle accident.

[0068] The method, device, equipment, storage medium and program product for constructing the above-mentioned risk prediction model, the method performs data expansion according to the distribution of the historical data of the vehicle to obtain sample data, and then inputs the sample data into the initial risk prediction model to perform risk prediction to obtain the prediction result, and finally adjusts the initial risk prediction model according to the prediction result to obtain the target risk prediction model. In the above-mentioned method, data expansion is performed on the historical data of the vehicle to obtain a large amount of sample data, and then the risk prediction model is optimized based on the large amount of sample data, which can avoid the problem of inaccurate prediction caused by insufficient sample data in traditional methods, thereby improving the accuracy of the risk prediction model. At the same time, data expansion according to the distribution of the historical data of the vehicle can improve the efficiency of model optimization to a certain extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 is an internal structure diagram of a computer device in one embodiment;

[0070] Figure 2 A schematic diagram of a process for constructing a risk prediction model in an embodiment;

[0071] Figure 3 A schematic diagram of a process for constructing a risk prediction model in another embodiment;

[0072] Figure 4 A schematic diagram of a process for constructing a risk prediction model in another embodiment;

[0073] Figure 5 A schematic diagram of a process for constructing a risk prediction model in another embodiment;

[0074] Figure 6 A schematic diagram of a process for constructing a risk prediction model in another embodiment;

[0075] Figure 7 A schematic diagram of a process for constructing a risk prediction model in another embodiment;

[0076] Figure 8 A schematic diagram of a risk prediction method in one embodiment;

[0077] Fig. 9 A schematic diagram of a risk prediction method in another embodiment;

[0078] Fig.10 is a structural block diagram of a device for constructing a risk prediction model in one embodiment;

[0079] Fig.11 4 is a structural block diagram of a risk prediction device in one embodiment. DETAILED DESCRIPTION

[0080] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0081] With the development of network information, big data analysis has the advantages of high efficiency, accuracy and real-time, and has been applied to many fields, especially in vehicle risk assessment. At present, a vehicle risk prediction model can usually be constructed based on algorithms such as neural networks or machine learning, which is used to analyze and predict the correlation between the vehicle and traffic accidents, so as to reduce the occurrence of major traffic accidents. However, the current risk prediction model has the problem of inaccuracy. The present application provides a method for constructing a risk prediction model and a risk prediction method, aiming to solve the above-mentioned technical problems. The following embodiments will specifically illustrate the method for constructing a risk prediction model and the risk prediction method described in the present application.

[0082] The method for constructing the risk prediction model provided in the embodiments of the present application can be applied to Figure 1 In the computer device shown in FIG. 1 , the computer device may be a terminal or a server, and its internal structure diagram may be as shown in FIG. Figure 1 As shown, the computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a method for constructing a risk prediction model and a risk prediction method are realized. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0083] Those skilled in the art will understand that Figure 1 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0084] In one embodiment, Figure 2 As shown in the figure, a method for constructing a risk prediction model is provided, and the method is applied to Figure 1 The computer device in the example is used to illustrate, including the following steps:

[0085] S201, perform data expansion according to the distribution of historical data of vehicles to obtain sample data.

[0086] The historical data of a vehicle includes at least one of basic vehicle information data, vehicle trajectory data, vehicle use and maintenance data, vehicle violation data, vehicle accident data, vehicle monitoring location data, data generated by real-time vehicle speed, vehicle failure data, dispatch vehicle data, driver fatigue monitoring data, positioning data, etc. Sample data is data expanded based on historical data.

[0087] In an embodiment of the present application, the computer device can obtain the historical data of various vehicles by means of network crawling, and specifically can obtain the historical data of the vehicle through a traffic management platform, a vehicle-mounted terminal, monitoring technology, GPS positioning technology or Beidou high-precision positioning technology. Then the historical data of the vehicle can be analyzed to determine the distribution of the historical data of the vehicle, and then a large amount of sample data can be generated according to the distribution of the historical data of the vehicle. Optionally, after obtaining the historical data of the vehicle, the computer device can first pre-process the historical data of the vehicle, and specifically can perform full data, incremental data extraction, data cleaning, denoising, de-duplication, normalization and other processing to obtain the pre-processed historical data of the vehicle, and then analyze the pre-processed historical data of the vehicle to determine the distribution of the historical data of the vehicle, and then generate a large amount of sample data according to the distribution of the historical data of the vehicle.

[0088] S202, inputting the sample data into the initial risk prediction model to perform risk prediction and obtain prediction results.

[0089] The initial risk prediction model may be a neural network model, an algorithm learning model, or other models. In this embodiment, a Monte Carlo simulation algorithm model is used. The prediction result is the risk value of the vehicle accident.

[0090] In the embodiment of the present application, after the computer device obtains a large amount of sample data based on the above steps, all the sample data can be uniformly input into the initial risk prediction model for risk prediction to obtain a large number of prediction results. Optionally, the large amount of sample data can be divided into multiple groups of sample data, and then each group of sample data is input into the initial risk prediction model for risk prediction to obtain the prediction results corresponding to each group of sample data.

[0091] S203, adjusting the initial risk prediction model according to the prediction result to obtain a target risk prediction model.

[0092] Among them, the target risk prediction model is used to predict the risk of vehicle accidents.

[0093] In an embodiment of the present application, after the computer device obtains the prediction result, the prediction result can be compared with the expected result. If the comparison result indicates that the prediction result meets the threshold requirement of the expected result, the initial risk prediction model is determined as the target risk prediction model. If the comparison result indicates that the prediction result does not meet the threshold requirement of the expected result, the initial risk prediction model is adjusted according to the prediction result until the prediction result meets the threshold requirement of the expected result.

[0094] The method for constructing a risk prediction model provided in an embodiment of the present application performs data expansion according to the distribution of the historical data of the vehicle to obtain sample data, and then inputs the sample data into the initial risk prediction model to perform risk prediction to obtain a prediction result, and finally adjusts the initial risk prediction model according to the prediction result to obtain a target risk prediction model. In the above method, data expansion is performed on the historical data of the vehicle to obtain a large amount of sample data, and then the risk prediction model is optimized based on the large amount of sample data, which can avoid the problem of inaccurate prediction caused by insufficient sample data in traditional methods, thereby improving the accuracy of the risk prediction model. At the same time, data expansion according to the distribution of the historical data of the vehicle can improve the efficiency of model optimization to a certain extent.

[0095] In one embodiment, a method for data expansion is also provided, such as Figure 3 As shown, the above step S201 of "expanding data according to the distribution of historical data of vehicles to obtain sample data" includes:

[0096] S301, classify historical data to obtain multiple types of historical data.

[0097] Among them, multiple types of historical data include vehicle road distribution data, vehicle work efficiency index data, vehicle utilization index data, vehicle maintenance cost data, vehicle violation rate index data, vehicle accident rate index data, etc.

[0098] In the embodiment of the present application, after the computer device obtains the historical data of the vehicle, the historical data of the vehicle can be input into a preset classification model for classification to obtain multiple types of historical data. Optionally, other classification techniques can be used to classify the historical data to obtain multiple types of historical data.

[0099] S302, performing data expansion according to the distribution of each type of historical data to obtain sample data.

[0100] In an embodiment of the present application, after the computer device obtains multiple types of historical data, it can perform statistical analysis on each type of historical data to determine the distribution of each type of historical data, and then perform data expansion based on the distribution of each type of historical data to obtain sample data.

[0101] Optionally, in one embodiment, Figure 4 As shown, the "expanding data according to the distribution of each type of historical data to obtain sample data" in the above step S302 includes:

[0102] S401, according to the distribution conditions of various types of historical data, determine the distribution relationship corresponding to each distribution condition.

[0103] In the embodiment of the present application, after obtaining the distribution of each type of historical data, the computer device can fit the distribution of each type of historical data to fit the distribution relationship corresponding to each distribution situation. Specifically, each type of historical data can be input into a preset curve fitting software to perform curve fitting to obtain the distribution relationship corresponding to each distribution situation.

[0104] S402, randomly generating data according to a preset number of samples and various distribution relationship equations to form multiple groups of sample data.

[0105] Each set of sample data includes all types of data.

[0106] In the embodiment of the present application, after obtaining the distribution relationship corresponding to each distribution situation, the computer device can randomly generate a preset number of sample data according to each distribution relationship. Specifically, a sample data of the corresponding type can be generated according to each type of distribution relationship, and a group of sample data containing each type can be obtained. Then, according to the number of times corresponding to the preset number of samples, the step of generating a sample data of the corresponding type according to each type of distribution relationship is repeated, and finally a preset number of sample data is generated. For example, the distribution relationship corresponding to type A is f (A), the distribution relationship corresponding to type B is f (B), and so on. Then, a sample data a1 can be randomly generated from f (A), and a sample data b1 can be randomly generated from f (B), so as to obtain a group of sample data (a1, b1). Repeating the above steps, multiple groups of sample data (a1, b1), (a2, b2), ..., (an, bn), etc. can be obtained.

[0107] Optionally, for each type of distribution relationship, a preset number of sample data may be generated to obtain multiple sample data corresponding to each type, and then the multiple sample data corresponding to each type may be randomly combined to obtain multiple groups of sample data. For example, the distribution relationship corresponding to type A is f(A), the distribution relationship corresponding to type B is f(B), and so on, and then multiple sample data a11, a12, ..., am may be randomly generated from f(A), and multiple sample data b11, b12, ..., bm may be randomly generated from f(B), and then (a11, a12, ..., am) and (b11, b12, ..., bm) may be randomly combined to obtain multiple groups of sample data (a1, b1), (a2, b2), ..., (an, bn), and so on.

[0108] The method described in the embodiments of the present application, by fitting the distribution relationship corresponding to each type of historical data and then expanding the data based on the distribution relationship, can not only obtain a large amount of sample data, but also make the generated sample data close to the real data, which can further improve the prediction accuracy of the model.

[0109] In one embodiment, the above prediction results include the predicted accident rate, and a specific implementation method of risk prediction is also provided, such as Figure 5 As shown, the above step S202 of "inputting the sample data into the initial risk prediction model to perform risk prediction and obtain a prediction result" includes:

[0110] S501, inputting sample data into an initial risk prediction model for risk prediction, and obtaining the risk value corresponding to each group of sample data.

[0111] In the embodiment of the present application, the computer device can determine the risk function in the initial risk prediction model in advance based on various types of historical data and the degree of influence of various types of historical data on the occurrence of accidents. For example, the risk function in the initial risk prediction model can be expressed by the relationship:

[0112]

[0113] Among them, A is the risk value, P is the vehicle accident rate index data, L is the vehicle work efficiency index data, Q is the vehicle road distribution data, and a, b, and c are the impact levels. It should be noted that the above risk function is composed of some types of historical data, and optionally, it can also be composed of all types of historical data, that is, it can be composed of vehicle road distribution data, vehicle work efficiency index data, vehicle utilization index data, vehicle maintenance cost data, vehicle violation rate index data, vehicle accident rate index data, etc.

[0114] After obtaining a large amount of sample data, the computer device can input the large amount of sample data into an initial risk prediction model and generate multiple risk values ​​corresponding to the sample data through a risk function.

[0115] S502, performing accident assessment according to the risk value corresponding to each group of sample data to obtain a predicted accident rate.

[0116] Among them, the prediction results include the predicted accident rate.

[0117] In the embodiment of the present application, after the computer device obtains multiple risk values ​​corresponding to the sample data based on the above steps, the multiple risk values ​​corresponding to the sample data can be input into a preset model or a preset functional relationship to calculate the predicted accident rate. As an optional method, the mean and standard deviation corresponding to each risk value can be calculated first, and then the predicted accident rate can be calculated by accident rate = mean + standard deviation * Z value, where the Z value is the quantile corresponding to the selected confidence level in the standard normal distribution, which is a preset value.

[0118] The method described in the embodiment of the present application predicts the accident rate of the vehicle through the risk value, and the initial risk prediction model can be adjusted based on the predicted accident rate, so that the adjusted model can have a higher prediction accuracy.

[0119] In one embodiment, a specific implementation method for adjusting the initial risk prediction model to obtain the target risk prediction model is also provided, such as Figure 6 As shown, the above step S203 of "adjusting the initial risk prediction model according to the prediction result to obtain the target risk prediction model" includes:

[0120] S601, verify the prediction result to obtain a verification result.

[0121] The verification result includes verification passed or verification failed.

[0122] In an embodiment of the present application, after obtaining the prediction result, the computer device can further obtain the expected result, and then verify the prediction result based on the prediction result and the expected result to obtain a verification result.

[0123] Optional, such as Figure 7 As shown, the "verifying the prediction result to obtain the verification result" in the above step S601 includes:

[0124] S6011, obtain part of the data from the historical data as verification data.

[0125] Among them, the verification data is used to verify whether the prediction results meet the preset requirements.

[0126] In the embodiment of the present application, the computer device can randomly obtain part of the historical data from the vehicle's historical data as verification data.

[0127] S6012, perform accident assessment on the verification data to obtain the actual accident rate.

[0128] In the embodiment of the present application, after the computer device obtains the verification data, it can obtain the actual accident rate by statistically analyzing the verification data. Specifically, it can first determine the abnormal data in the verification data, and then determine the actual accident rate based on the abnormal data.

[0129] S6013, verify the prediction result according to the actual accident rate to obtain a verification result.

[0130] In an embodiment of the present application, after the computer device obtains the predicted accident rate and the actual accident rate, it can determine whether the predicted accident rate meets the threshold requirement of the actual accident rate. If so, it is determined that the verification result indicates that the verification is passed, and step S602 is executed. If not, it is determined that the verification result indicates that the verification is not passed, and step S603 is executed.

[0131] S602, determining the initial risk prediction model as the target risk prediction model.

[0132] In the embodiment of the present application, if the verification passes, the initial risk prediction model is determined as the target risk prediction model.

[0133] S603, adjusting the parameters in the initial risk prediction model to obtain a target risk prediction model.

[0134] In an embodiment of the present application, if the verification fails, the parameters of the risk function in the initial risk prediction model are adjusted, and then the process returns to step S501 until the predicted accident rate meets the threshold requirement of the actual accident rate, and the target risk prediction model can be obtained.

[0135] The method described in the embodiments of the present application can obtain a target risk prediction model that can accurately predict by verifying the prediction results and then adjusting the risk function in the initial risk prediction model based on the verification results.

[0136] In one embodiment, Figure 8 As shown, a risk prediction method is also provided, the method comprising:

[0137] S701, acquiring accident data of multiple accident vehicles.

[0138] In an embodiment of the present application, when it is necessary to analyze the associated factors of a vehicle accident, the computer device can obtain accident data of multiple accident vehicles in a preset time period.

[0139] S702, input each accident data into the target risk prediction model for risk prediction, and obtain the risk value corresponding to each accident data.

[0140] In an embodiment of the present application, a computer device can obtain a target risk prediction model determined based on the above embodiment, and then after obtaining the accident data of multiple accident vehicles, the accident data of multiple accident vehicles can be input into the target risk prediction model, and the risk value corresponding to each accident data can be calculated through the risk function in the target risk prediction model.

[0141] S703: Analyze the risk value corresponding to each accident data and the corresponding accident data to determine the correlation factors of the vehicle accident.

[0142] Among them, the correlation factors of vehicle accidents are determined by various types of accident data.

[0143] In an embodiment of the present application, after obtaining each accident data and the risk value corresponding to each accident data, the computer device can perform a sensitivity analysis on the risk value corresponding to each accident data to determine the associated factors of the vehicle accident. Specifically, each type of accident data can be used as a variable, and the influence of each variable on the accident risk value can be analyzed through the control variable method, thereby determining the associated factors of the vehicle accident.

[0144] The method described in the embodiment of the present application, since the target risk prediction model can accurately predict the risk value corresponding to the accident data of the accident vehicle, the influencing factors of the vehicle accident can be accurately analyzed based on the risk value corresponding to each accident data.

[0145] Based on all the above embodiments, a risk prediction method is also provided. Fig. 9 As shown, the method includes:

[0146] S801, classify historical data to obtain multiple types of historical data.

[0147] S802: Determine distribution relationship expressions corresponding to various distribution conditions according to the distribution conditions of various types of historical data.

[0148] S803, randomly generating data according to the preset sample quantity and each distribution relationship to form multiple groups of sample data, wherein each group of sample data includes all types of data.

[0149] S804, inputting the sample data into the initial risk prediction model for risk prediction, and obtaining the risk value corresponding to each group of sample data.

[0150] S805, performing accident assessment according to the risk value corresponding to each group of sample data to obtain a predicted accident rate.

[0151] S806, verify the prediction result to obtain a verification result. If the verification result indicates that the verification is passed, execute S807; if the verification result indicates that the verification is not passed, execute S808.

[0152] S807, determining the initial risk prediction model as the target risk prediction model.

[0153] S808, adjusting the parameters in the initial risk prediction model, and returning to execute S804 to obtain the target risk prediction model.

[0154] S809, acquiring accident data of multiple accident vehicles.

[0155] S810, input each accident data into the target risk prediction model for risk prediction, and obtain the risk value corresponding to each accident data.

[0156] S811, analyzing the risk value corresponding to each accident data and the corresponding accident data to determine the correlation factors of the vehicle accident.

[0157] The method described in the embodiments of the present application can not only analyze the degree of influence of influencing factors on risk prediction and judgment, but also scientifically evaluate the accident rate of risk events, fully consider the joint influence and joint action of multiple influencing factors, and improve the prediction accuracy.

[0158] The methods described in the above steps are all described in the above embodiments. Please refer to the above description for details and will not be repeated here.

[0159] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0160] Based on the same inventive concept, the embodiment of the present application also provides a method for constructing a risk prediction model and a device for constructing a risk prediction model and a risk prediction device for implementing the risk prediction model involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the embodiments of the device for constructing one or more risk prediction models and the risk prediction device provided below can be found in the above limitations on the device for constructing a risk prediction model and the risk prediction method, and will not be repeated here.

[0161] In one embodiment, Fig.10 As shown, a device for constructing a risk prediction model is provided, comprising:

[0162] The expansion module 10 is used to expand the data according to the distribution of the historical data of the vehicle to obtain sample data.

[0163] The first prediction module 11 is used to input sample data into the initial risk prediction model to perform risk prediction and obtain prediction results.

[0164] The adjustment module 12 is used to adjust the initial risk prediction model according to the prediction results to obtain a target risk prediction model.

[0165] In one embodiment, the expansion module 10 includes:

[0166] The classification unit is used to classify the historical data to obtain multiple types of historical data.

[0167] The expansion unit is used to expand the data according to the distribution of various types of historical data to obtain sample data.

[0168] In one embodiment, the expansion unit comprises:

[0169] The determination subunit is used to determine the distribution relationship corresponding to each distribution situation according to the distribution situation of each type of historical data.

[0170] The generating subunit is used to randomly generate data according to a preset number of samples and various distribution relationship equations to form multiple groups of sample data; each group of sample data includes all types of data.

[0171] In one embodiment, the first prediction module 11 includes:

[0172] The prediction unit is used to input the sample data into the initial risk prediction model for risk prediction to obtain the risk value corresponding to each group of sample data.

[0173] The evaluation unit is used to perform accident evaluation according to the risk value corresponding to each group of sample data to obtain the predicted accident rate.

[0174] In one embodiment, the adjustment module 12 includes:

[0175] The verification unit is used to verify the prediction result and obtain the verification result.

[0176] The first determination unit is used to determine the initial risk prediction model as the target risk prediction model if the verification result indicates that the verification is passed.

[0177] The second determination unit is used to adjust the parameters in the initial risk prediction model to obtain a target risk prediction model if the verification result indicates that the verification has not passed.

[0178] In one embodiment, the verification unit includes:

[0179] The acquisition subunit is used to acquire part of the data from the historical data as verification data.

[0180] The evaluation subunit is used to perform accident evaluation on the verification data to obtain the actual accident rate.

[0181] The verification subunit is used to verify the prediction result according to the actual accident rate to obtain the verification result.

[0182] In one embodiment, Fig.11 As shown, a risk prediction device is also provided, comprising:

[0183] The acquisition module 20 is used to acquire accident data of multiple accident vehicles.

[0184] The second prediction module 21 is used to input each accident data into the target risk prediction model for risk prediction to obtain the risk value corresponding to each accident data.

[0185] The analysis module 22 is used to analyze the risk value corresponding to each accident data and the corresponding accident data to determine the correlation factors of the vehicle accident.

[0186] The risk prediction model construction device and each module in the risk prediction device can be implemented in whole or in part by software, hardware and a combination thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0187] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0188] Perform data expansion based on the distribution of historical data of vehicles to obtain sample data;

[0189] Input the sample data into the initial risk prediction model to perform risk prediction and obtain the prediction results;

[0190] The initial risk prediction model is adjusted according to the prediction results to obtain the target risk prediction model.

[0191] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0192] Classify historical data to obtain multiple types of historical data;

[0193] Data expansion is performed based on the distribution of various types of historical data to obtain sample data.

[0194] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0195] According to the distribution of various types of historical data, the distribution relationship corresponding to each distribution situation is determined;

[0196] The data randomly generated according to the preset sample quantity and various distribution relationship formulas constitute multiple groups of sample data; each group of sample data includes all types of data.

[0197] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0198] Input the sample data into the initial risk prediction model to perform risk prediction, and obtain the risk value corresponding to each group of sample data;

[0199] Accident assessment is performed based on the risk value corresponding to each group of sample data to obtain the predicted accident rate.

[0200] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0201] Verify the prediction results to obtain verification results;

[0202] If the verification result indicates that the verification is passed, the initial risk prediction model is determined as the target risk prediction model;

[0203] If the verification result indicates that the verification has not passed, the parameters in the initial risk prediction model are adjusted to obtain the target risk prediction model.

[0204] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0205] Get some data from historical data as verification data;

[0206] Conduct accident assessment on the verification data to obtain the actual accident rate;

[0207] The prediction results are verified according to the actual accident rate to obtain the verification results.

[0208] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0209] Obtain accident data of multiple accident vehicles;

[0210] Input each accident data into the target risk prediction model for risk prediction, and obtain the risk value corresponding to each accident data;

[0211] Analyze the risk value and the corresponding accident data corresponding to each accident data to determine the associated factors of the vehicle accident.

[0212] The above embodiment provides a computer device, whose implementation principle and technical effect are similar to those of the above method embodiment, and will not be repeated here.

[0213] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0214] Perform data expansion based on the distribution of historical data of vehicles to obtain sample data;

[0215] Input the sample data into the initial risk prediction model to perform risk prediction and obtain the prediction results;

[0216] The initial risk prediction model is adjusted according to the prediction results to obtain the target risk prediction model.

[0217] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0218] Classify historical data to obtain multiple types of historical data;

[0219] Data expansion is performed based on the distribution of various types of historical data to obtain sample data.

[0220] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0221] According to the distribution of various types of historical data, the distribution relationship corresponding to each distribution situation is determined;

[0222] The data randomly generated according to the preset sample quantity and various distribution relationship formulas constitute multiple groups of sample data; each group of sample data includes all types of data.

[0223] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0224] Input the sample data into the initial risk prediction model to perform risk prediction, and obtain the risk value corresponding to each group of sample data;

[0225] Accident assessment is performed based on the risk value corresponding to each group of sample data to obtain the predicted accident rate.

[0226] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0227] Verify the prediction results to obtain verification results;

[0228] If the verification result indicates that the verification is passed, the initial risk prediction model is determined as the target risk prediction model;

[0229] If the verification result indicates that the verification has not passed, the parameters in the initial risk prediction model are adjusted to obtain the target risk prediction model.

[0230] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0231] Get some data from historical data as verification data;

[0232] Conduct accident assessment on the verification data to obtain the actual accident rate;

[0233] The prediction results are verified according to the actual accident rate to obtain the verification results.

[0234] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0235] Obtain accident data of multiple accident vehicles;

[0236] Input each accident data into the target risk prediction model for risk prediction, and obtain the risk value corresponding to each accident data;

[0237] Analyze the risk value and the corresponding accident data corresponding to each accident data to determine the associated factors of the vehicle accident.

[0238] The above embodiment provides a computer-readable storage medium, whose implementation principle and technical effect are similar to those of the above method embodiment, and will not be repeated here.

[0239] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0240] Perform data expansion based on the distribution of historical data of vehicles to obtain sample data;

[0241] Input the sample data into the initial risk prediction model to perform risk prediction and obtain the prediction results;

[0242] The initial risk prediction model is adjusted according to the prediction results to obtain the target risk prediction model.

[0243] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0244] Classify historical data to obtain multiple types of historical data;

[0245] Data expansion is performed based on the distribution of various types of historical data to obtain sample data.

[0246] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0247] According to the distribution of various types of historical data, the distribution relationship corresponding to each distribution situation is determined;

[0248] The data randomly generated according to the preset sample quantity and various distribution relationship formulas constitute multiple groups of sample data; each group of sample data includes all types of data.

[0249] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0250] Input the sample data into the initial risk prediction model to perform risk prediction, and obtain the risk value corresponding to each group of sample data;

[0251] Accident assessment is performed based on the risk value corresponding to each group of sample data to obtain the predicted accident rate.

[0252] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0253] Verify the prediction results to obtain verification results;

[0254] If the verification result indicates that the verification is passed, the initial risk prediction model is determined as the target risk prediction model;

[0255] If the verification result indicates that the verification has not passed, the parameters in the initial risk prediction model are adjusted to obtain the target risk prediction model.

[0256] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0257] Get some data from historical data as verification data;

[0258] Conduct accident assessment on the verification data to obtain the actual accident rate;

[0259] The prediction results are verified according to the actual accident rate to obtain the verification results.

[0260] In one embodiment, a computer program product is also provided, including a computer program, which, when executed by a processor, implements the following steps:

[0261] Obtain accident data of multiple accident vehicles;

[0262] Input each accident data into the target risk prediction model for risk prediction, and obtain the risk value corresponding to each accident data;

[0263] Analyze the risk value and the corresponding accident data corresponding to each accident data to determine the associated factors of the vehicle accident.

[0264] The above embodiment provides a computer program product, whose implementation principle and technical effect are similar to those of the above method embodiment, and will not be repeated here.

[0265] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0266] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0267] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for constructing a risk prediction model, characterized in that: The method comprises: Perform data expansion based on the distribution of historical data of vehicles to obtain sample data; Inputting the sample data into an initial risk prediction model to perform risk prediction and obtain a prediction result; The initial risk prediction model is adjusted according to the prediction result to obtain a target risk prediction model.

2. The method according to claim 1, characterized in that: The data expansion is performed according to the distribution of the historical data of the vehicle to obtain sample data, including: Classifying the historical data to obtain multiple types of historical data; Data expansion is performed according to the distribution of each type of historical data to obtain the sample data.

3. The method according to claim 2, characterized in that The data expansion is performed according to the distribution of the historical data of each type to obtain the sample data, including: According to the distribution of the historical data of each type, determining the distribution relationship corresponding to each distribution situation; The data randomly generated according to the preset sample quantity and each of the distribution relationship formulas constitute multiple groups of sample data; each group of the sample data includes all types of data.

4. The method according to claim 1, characterized in that: The prediction result includes a predicted accident rate, and the sample data is input into an initial risk prediction model to perform risk prediction to obtain a prediction result, including: Inputting the sample data into the initial risk prediction model for risk prediction to obtain the risk value corresponding to each group of sample data; Accident assessment is performed according to the risk values ​​corresponding to the groups of sample data to obtain the predicted accident rate.

5. The method according to claim 1, characterized in that The step of adjusting the initial risk prediction model according to the prediction result to obtain a target risk prediction model includes: Verifying the prediction result to obtain a verification result; If the verification result indicates that the verification is passed, the initial risk prediction model is determined as the target risk prediction model; If the verification result indicates that the verification has not passed, the parameters in the initial risk prediction model are adjusted to obtain a target risk prediction model.

6. The method according to claim 5, characterized in that The step of verifying the prediction result to obtain a verification result includes: Acquire part of the historical data as verification data; Performing accident assessment on the verification data to obtain an actual accident rate; The prediction result is verified according to the actual accident rate to obtain a verification result.

7. A risk prediction method, characterized in that: The method comprises: Obtain accident data of multiple accident vehicles; Inputting each of the accident data into a target risk prediction model for risk prediction to obtain a risk value corresponding to each of the accident data; An analysis is performed based on the risk value corresponding to each of the accident data and the corresponding accident data to determine the associated factors of the vehicle accident.

8. A device for constructing a risk prediction model, characterized in that: The device comprises: An expansion module is used to expand data according to the distribution of historical data of vehicles to obtain sample data; A first prediction module is used to input the sample data into an initial risk prediction model to perform risk prediction and obtain a prediction result; An adjustment module is used to adjust the initial risk prediction model according to the prediction result to obtain a target risk prediction model.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images