CN109784586B - Prediction method and system for danger emergence condition of vehicle danger - Google Patents
Prediction method and system for danger emergence condition of vehicle danger Download PDFInfo
- Publication number
- CN109784586B CN109784586B CN201910173365.7A CN201910173365A CN109784586B CN 109784586 B CN109784586 B CN 109784586B CN 201910173365 A CN201910173365 A CN 201910173365A CN 109784586 B CN109784586 B CN 109784586B
- Authority
- CN
- China
- Prior art keywords
- risk
- policy data
- historical
- vehicle
- prediction model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Abstract
The invention discloses a method and a system for predicting the danger situation of a vehicle danger, wherein the method for predicting the danger situation of the vehicle danger comprises the following steps: acquiring a plurality of pieces of first historical policy data and corresponding historical risk-giving results; calculating the weight of each first historical policy data according to the first historical policy data and the historical risk finding result; inputting the weight and the first historical policy data into a classification model for training to obtain a car insurance risk prediction model, wherein the output parameters of the car insurance risk prediction model comprise predicted risk result; and inputting the policy data to be predicted into a car insurance risk prediction model to predict whether the policy data to be predicted is a predicted risk result of risk. Compared with the traditional vehicle risk prediction model, the method provided by the invention has the advantage that the lifting degree is greatly improved.
Description
Technical Field
The invention relates to the field of data processing, in particular to a method and a system for predicting the emergence condition of a vehicle danger.
Background
Conventional car insurance prediction models generally use the annual number of risk, risk amount or odds of each car as target variables, and fit the target values by using other factors (car type, car price, region, risk, driving characteristics before purchasing insurance, historical risk situation, etc.) through a supervised class algorithm. However, the dangerous situation of the vehicle has the necessity caused by factors such as driving habit, road condition and the like, and is accompanied by contingency. This is because even with the same driving behavior, the final risk number is not likely to be exactly the same; even if the number of hazards is the same, the impact location and damage of the accident vehicle will be different due to different types of accidents, resulting in different payouts and payouts.
Therefore, the target selection mode in the traditional vehicle risk modeling process leads the accidental factors in the target variables to lead the generation of the model, the model can not locate the real risk factors from the input characteristics, the problems of easy overfitting and the like are caused, and finally, the inaccurate prediction result of the vehicle risk prediction model in the practical application process after the vehicle risk prediction model is on line is caused.
Disclosure of Invention
The invention aims to overcome the defect that a vehicle insurance prediction model in the prior art is inaccurate in prediction, and provides a method and a system for predicting the emergence condition of vehicle insurance.
The invention solves the technical problems through the following technical scheme:
a method for predicting an emergency situation of a vehicle emergency, the method comprising:
acquiring a plurality of pieces of first historical policy data and corresponding historical risk-giving results;
calculating the weight of each first historical policy data according to the first historical policy data and the historical risk-giving result;
inputting the weight and the first historical policy data into a classification model for training to obtain a car insurance risk prediction model, wherein output parameters of the car insurance risk prediction model comprise a predicted risk result;
and inputting the policy data to be predicted into the car insurance risk prediction model to predict whether the policy data to be predicted is a predicted risk result of risk.
Preferably, the output parameters of the vehicle risk prediction model further comprise scoring values; the step of calculating the weight of each first historical policy data according to the first historical policy data and the historical risk-solving result comprises the following steps:
calculating and obtaining the weight by adopting at least two preset calculation methods;
the step of inputting the weight and the first historical policy data into a classification model for training to obtain a car insurance risk prediction model comprises the following steps:
respectively taking the weight obtained by calculation of each preset calculation method and the first historical policy data as input to obtain respectively corresponding vehicle risk prediction models as vehicle risk prediction models to be selected;
respectively inputting a plurality of pieces of second historical policy data into a to-be-selected vehicle risk prediction model to respectively obtain a corresponding prediction risk result and a scoring value; calculating the lifting degree of the vehicle insurance risk prediction model according to the historical insurance risk results corresponding to the scoring value and the second historical policy data, and selecting the model with the highest lifting degree in the vehicle insurance risk prediction model to be selected as the final vehicle insurance risk prediction model
Preferably, the method comprises the steps of,
the preset calculation method comprises the following steps:
wherein weight1 is the weight, claimcnt i For the number of danger, r i For the annual proportion, α is a limiting parameter, within a preset threshold range;
and/or the number of the groups of groups,
wherein weight2 is the weight, claimcnt i For the number of danger, r i Beta is a parameter of evolution for annual proportion;
and/or the number of the groups of groups,
wherein weight3 is the weight, claimcnt i For the number of danger, r i For annual proportion, β is the parameter of the formulation.
Preferably, after the step of inputting the policy data to be predicted into the vehicle risk prediction model to predict whether the policy data to be predicted is a predicted risk result of risk, the prediction method of the risk situation of the vehicle risk further includes:
and judging whether the service time of the car insurance risk prediction model exceeds a preset period, and if so, returning to the step of acquiring a plurality of pieces of first historical policy data and corresponding historical risk results.
Preferably, the first historical policy data and/or the second historical policy data include owner information, vehicle information and insurance information.
The prediction system of the dangerous situation of the vehicle insurance comprises an acquisition module, a weight calculation module, a training module and a prediction module;
the acquisition module is used for acquiring a plurality of pieces of first historical policy data and corresponding historical risk-giving results;
the weight calculation module is used for calculating the weight of each first historical policy data according to the first historical policy data and the historical risk-giving result;
the training module is used for inputting the weight and the first historical policy data into a classification model for training to obtain a car insurance risk prediction model, and output parameters of the car insurance risk prediction model comprise a predicted risk result;
the prediction module is used for inputting the policy data to be predicted into the car insurance risk prediction model so as to predict the predicted risk result of whether the policy data to be predicted is in risk or not.
Preferably, the output parameters of the vehicle risk prediction model further comprise scoring values;
the weight calculation module is also used for calculating and obtaining the weight by adopting at least two preset calculation methods;
the training module is further used for taking the weights obtained by calculation of each preset calculation method and the first historical policy data as input to obtain respectively corresponding car insurance risk prediction models as car insurance risk prediction models to be selected;
the training module is also used for respectively inputting a plurality of second historical policy data into the to-be-selected vehicle risk prediction model to respectively obtain a corresponding prediction risk result and a scoring value; calculating the lifting degree of the vehicle insurance risk prediction model according to the scoring value and the historical risk result corresponding to the second historical policy data, and selecting the model with the highest lifting degree in the vehicle insurance risk prediction models to be selected as the final vehicle insurance risk prediction model
Preferably, the method comprises the steps of,
the preset calculation method comprises the following steps:
wherein weight1 is the weight, claimcnt i For the number of danger, r i For the annual proportion, α is a limiting parameter, within a preset threshold range;
and/or the number of the groups of groups,
wherein weight2 is the weight, claimcnt i For the number of danger, r i Beta is a parameter of evolution for annual proportion;
and/or the number of the groups of groups,
wherein weight3 is the weight, claimcnt i For the number of danger, r i For annual proportion, β is the parameter of the formulation.
Preferably, the prediction system of the risk situation of the vehicle risk further comprises an updating module, wherein the updating module is used for judging whether the service time of the risk prediction model exceeds a preset period, and if yes, the collecting module is called.
Preferably, the first historical policy data and/or the second historical policy data include owner information, vehicle information and insurance information.
The invention has the positive progress effects that:
according to the invention, a mode that the traditional vehicle insurance prediction model predicts continuous values such as the number of times of insurance, the amount of money, the odds ratio and the like is changed into a two-class mode of predicting whether to carry out insurance, the vehicle insurance prediction model obtained by training is obtained by taking historical policy data and corresponding weights as inputs, and the promotion degree of the vehicle insurance prediction model is greatly promoted compared with that of the traditional method prediction model.
Drawings
Fig. 1 is a flow chart of a method for predicting the risk of a vehicle risk according to embodiment 1 of the present invention.
Fig. 2 is a flowchart illustrating step 103 of the method for predicting the risk of vehicle risk according to embodiment 2 of the present invention.
Fig. 3 is a flow chart of a method for predicting the risk of vehicle risk according to embodiment 2 of the present invention.
Fig. 4 is a block diagram of a prediction system of a risk occurrence of a vehicle risk according to embodiment 3 of the present invention.
Fig. 5 is a block diagram of a prediction system of a risk occurrence of a vehicle risk according to embodiment 4 of the present invention.
Detailed Description
The invention is further illustrated by means of the following examples, which are not intended to limit the scope of the invention.
Example 1
The embodiment provides a method for predicting the risk of a vehicle risk, as shown in fig. 1, including:
step 101, acquiring a plurality of pieces of first history policy data and corresponding history risk results.
The risk-out results typically include number of risk-out, amount paid, rate paid, and the like.
And 102, calculating the weight of each first historical policy data according to the first historical policy data and the historical risk-giving result.
And step 103, inputting the weight and the first historical policy data into a classification model for training to obtain a car insurance risk prediction model, wherein the output parameters of the car insurance risk prediction model comprise a predicted risk result.
In this embodiment, each training data corresponds to a policy when model training is performed, and the first policy corresponds to a situation of a certain vehicle in a certain year if the first policy includes the first policy. If the vehicle is in danger in the current year, setting the target variable of the danger result to be 1 no matter the number of times and the amount of money, namely, the danger is in danger; if no danger is generated, the value is set to 0, namely, the danger is not generated. I.e. if the target variable is a risk, expressed as 0/1.
And 104, inputting the policy data to be predicted into a car insurance risk prediction model to predict whether the policy data to be predicted is a predicted risk result of risk.
According to the method, the device and the system, the traditional prediction model is converted into the two types of prediction of whether to risk by adopting the mode of predicting continuous values such as the number of times of risk, the amount of money and the like, the prediction results that different risk factors such as vehicles and driving can necessarily cause different numbers of risk and amounts of money in the original mode are solved, the mode that the target of prediction of the traditional vehicle risk prediction model is the continuous values such as the number of times of risk, the amount of money, the odds of pays and the like is changed into the two types of prediction of whether to risk, the vehicle risk prediction model obtained by training is input by utilizing historical policy data and corresponding weights, and the improvement of the vehicle risk prediction model is greatly improved compared with the improvement degree of the traditional method prediction model.
Example 2
Compared with the embodiment 1, the prediction method for the risk emergence condition of the vehicle risk provided by the embodiment is different in that the output parameters of the vehicle risk emergence prediction model further comprise scoring values;
step 102 comprises:
and calculating and obtaining the weight by adopting at least two preset calculation methods.
The result of the insurance policy usually includes the number of times of the insurance, the amount of the insurance, the odds, the annual proportion of the odds, etc.
In this embodiment, the selected preset calculation method is as follows:
wherein weight1 is the weight, claimcnt i For the number of danger, r i For the annual proportion, α is a limiting parameter, within a preset threshold range;
and/or the number of the groups of groups,
wherein weight2 is the weight, claimcnt i For the number of danger, r i Beta is a parameter of evolution for annual proportion;
and/or the number of the groups of groups,
wherein weight3 is the weight, claimcnt i For the number of danger, r i For annual proportion, β is the parameter of the formulation.
Correspondingly, step 103 includes:
1031. and respectively taking the weight obtained by calculation of each preset calculation method and the first security data as input to obtain respectively corresponding vehicle risk prediction models as vehicle risk prediction models to be selected.
Typically, policy data includes owner information, vehicle information, insurance information.
1032. And respectively inputting a plurality of pieces of second historical policy data into to-be-selected vehicle insurance risk prediction models to respectively obtain corresponding prediction risk results and scoring values, and calculating the lifting degree of the vehicle insurance risk prediction models according to the scoring values and the historical risk results corresponding to the second historical policy data, wherein the model with the highest lifting degree in the to-be-selected vehicle insurance risk prediction models is used as a final vehicle insurance risk prediction model.
The prediction method of the dangerous situation of the vehicle danger further comprises the following steps:
and 105, judging whether the service time of the vehicle risk prediction model exceeds a preset period, and if so, returning to the step 101.
After the car insurance prediction model is used on line, the sample updating training can be regularly carried out on the car insurance prediction model due to the change of various external environments so as to optimize the prediction accuracy of the car insurance prediction model at any time.
In this embodiment, several historical policy data are collected, and these historical policy data are divided into three parts, one part is used as a training data set, namely a first historical data, the other part is used as a verification data set, namely a second historical data, and the other part is used as a test data set for testing the lifting effect of the vehicle risk prediction model, and the three parts of data can be randomly divided in a ratio of 5:3:2.
The two classification algorithm of this embodiment may select two classification models such as random forest, GBDT, deep neural network, etc., take whether to take the risk (0/1) as its target variable, utilize the weights calculated by the above three calculation methods to respectively take as the input weight parameters of the model, train on the training dataset, and respectively obtain the corresponding car risk prediction model, then score on the verification set, i.e. after the verification dataset is ranked from high to low according to the model score, equally divide into 10 shares, calculate the average value of the risk number, amount or odds in the corresponding historical risk result in each share, and divide the highest score in each segment by the lowest score to obtain the corresponding value, i.e. respectively obtain the predicted lifting degree of risk number, amount or odds, according to practical application, the car risk prediction model corresponding to the appropriate predicted lifting degree is selected to be used as the optimal car risk prediction model for online use.
Compared with the traditional prediction model, the prediction accuracy improvement effect of the vehicle risk prediction model obtained through training by the method of the embodiment can be evaluated by using the improvement degree, and the improvement degree obtained by the traditional method is about 2, and the improvement degree of the method of the embodiment is about 3 and is improved by 50%.
Example 3
The present embodiment provides a prediction system for the risk of vehicle risk, as shown in fig. 4, which includes an acquisition module 201, a weight calculation module 202, a training module 203, and a prediction module 204.
The acquisition module 201 is configured to acquire a plurality of pieces of first historical policy data and corresponding historical risk-giving results; the risk-out results typically include number of risk-out, amount paid, rate paid, and the like.
The weight calculation module 202 is configured to calculate a weight of each first historical policy data according to the first historical policy data and the historical risk-of-occurrence result;
the training module 203 is configured to input the weight and the first historical policy data into a classification model to train, so as to obtain a vehicle risk prediction model, where output parameters of the vehicle risk prediction model include a predicted risk result;
in this embodiment, each training data corresponds to a policy when model training is performed, and the first policy corresponds to a situation of a certain vehicle in a certain year if the first policy includes the first policy. If the vehicle is in danger in the current year, setting the target variable of the danger result to be 1 no matter the number of times and the amount of money, namely, the danger is in danger; if no danger is generated, the value is set to 0, namely, the danger is not generated. I.e. if the target variable is a risk, expressed as 0/1.
The prediction module 204 is configured to input the policy data to be predicted into the vehicle risk prediction model to predict whether the policy data to be predicted is a predicted risk result of risk.
According to the method, the device and the system, the traditional prediction model is converted into the two types of prediction of whether to risk by adopting the mode of predicting continuous values such as the number of times of risk, the amount of money and the like, the prediction results that different risk factors such as vehicles and driving can necessarily cause different numbers of risk and amounts of money in the original mode are solved, the mode that the target of prediction of the traditional vehicle risk prediction model is the continuous values such as the number of times of risk, the amount of money, the odds of pays and the like is changed into the two types of prediction of whether to risk, the vehicle risk prediction model obtained by training is input by utilizing historical policy data and corresponding weights, and the improvement of the vehicle risk prediction model is greatly improved compared with the improvement degree of the traditional method prediction model.
Example 4
The present embodiment provides a difference between the present embodiment and embodiment 3, in that the output parameters of the vehicle risk prediction model further include a scoring value.
The weight calculation module 202 is further configured to calculate and obtain weights by at least two preset calculation methods.
The result of the insurance policy usually includes the number of times of the insurance, the amount of the insurance, the odds, the annual proportion of the odds, etc.
In this embodiment, the selected preset calculation method is as follows:
wherein weight1 is the weight, claimcnt i For the number of danger, r i For the annual proportion, α is a limiting parameter, within a preset threshold range;
and/or the number of the groups of groups,
wherein weight2 is the weight, claimcnt i For the number of danger, r i Beta is a parameter of evolution for annual proportion;
and/or the number of the groups of groups,
wherein weight3 is the weight, claimcnt i For the number of danger, r i For annual proportion, β is the parameter of the formulation.
The training module 203 is further configured to input weights obtained by calculation by each preset calculation method and the first historical policy data, so as to obtain respectively corresponding vehicle risk prediction models as vehicle risk prediction models to be selected;
the training module 203 is further configured to input a plurality of pieces of second historical policy data to the to-be-selected vehicle risk prediction models respectively to obtain corresponding prediction risk results and scoring values, calculate a lifting degree of the vehicle risk prediction models according to the scoring values and the historical risk results corresponding to the second historical policy data, and select a model with the highest lifting degree from the to-be-selected vehicle risk prediction models as a final vehicle risk prediction model.
The prediction system of the danger situation of the vehicle danger further comprises an updating module 205, wherein the updating module is used for judging whether the service time of the danger prediction model exceeds a preset period, and if yes, the acquisition module is called.
After the car insurance prediction model is used on line, the sample updating training can be regularly carried out on the car insurance prediction model due to the change of various external environments so as to optimize the prediction accuracy of the car insurance prediction model at any time.
In this embodiment, several historical policy data are collected, and these historical policy data are divided into three parts, one part is used as a training data set, namely a first historical data, the other part is used as a verification data set, namely a second historical data, and the other part is used as a test data set for testing the lifting effect of the vehicle risk prediction model, and the three parts of data can be randomly divided in a ratio of 5:3:2.
The two classification algorithm of this embodiment may select two classification models such as random forest, GBDT, deep neural network, etc., take whether to take the risk (0/1) as its target variable, utilize the weights calculated by the above three calculation methods to respectively take as the input weight parameters of the model, train on the training dataset, and respectively obtain the corresponding car risk prediction model, then score on the verification set, i.e. after the verification dataset is ranked from high to low according to the model score, equally divide into 10 shares, calculate the average value of the risk number, amount or odds in the corresponding historical risk result in each share, and divide the highest score in each segment by the lowest score to obtain the corresponding value, i.e. respectively obtain the predicted lifting degree of risk number, amount or odds, according to practical application, the car risk prediction model corresponding to the appropriate predicted lifting degree is selected to be used as the optimal car risk prediction model for online use.
Compared with the traditional prediction model, the prediction accuracy and the lifting effect of the vehicle risk prediction model obtained through the system training of the embodiment can be evaluated by using the lifting degree, and the lifting degree obtained by the traditional method is about 2, and the lifting degree of the method of the embodiment is about 3 and is 50%.
While specific embodiments of the invention have been described above, it will be appreciated by those skilled in the art that these are by way of example only, and the scope of the invention is defined by the appended claims. Various changes and modifications to these embodiments may be made by those skilled in the art without departing from the principles and spirit of the invention, but such changes and modifications fall within the scope of the invention.
Claims (6)
1. The method for predicting the dangerous situation of the vehicle insurance is characterized by comprising the following steps:
acquiring a plurality of pieces of first historical policy data and corresponding historical risk-giving results;
calculating the weight of each first historical policy data according to the first historical policy data and the historical risk-giving result;
inputting the weight and the first historical policy data into a classification model for training to obtain a car insurance risk prediction model, wherein output parameters of the car insurance risk prediction model comprise a predicted risk result;
inputting the policy data to be predicted into the car insurance risk prediction model to predict whether the policy data to be predicted is a predicted risk result of risk;
the output parameters of the vehicle risk prediction model also comprise scoring values;
the step of calculating the weight of each first historical policy data according to the first historical policy data and the historical risk-solving result comprises the following steps:
calculating and obtaining the weight by adopting at least two preset calculation methods;
the step of inputting the weight and the first historical policy data into a classification model for training to obtain a car insurance risk prediction model comprises the following steps:
respectively taking the weight obtained by calculation of each preset calculation method and the first historical policy data as input to obtain respectively corresponding vehicle risk prediction models as vehicle risk prediction models to be selected;
respectively inputting a plurality of pieces of second historical policy data into the to-be-selected vehicle risk prediction model to respectively obtain a corresponding prediction risk result and a scoring value; calculating the lifting degree of the vehicle insurance risk prediction model according to the scoring value and the historical risk result corresponding to the second historical policy data, and selecting the model with the highest lifting degree in the vehicle insurance risk prediction models to be selected as a final vehicle insurance risk prediction model;
the promotion degree is used for representing model evaluation parameters related to the number of risk, amount or odds;
the preset calculation method comprises the following steps:
wherein weight1 is the weight, claimcnt i For the number of danger, r i For the annual proportion, α is a limiting parameter, within a preset threshold range;
and/or the number of the groups of groups,
wherein weight2 is the weight, claimcnt i For the number of danger, r i Beta is a parameter of evolution for annual proportion;
and/or the number of the groups of groups,
wherein weight3 is the weight, claimcnt i For the number of danger, r i For annual proportion, β is the parameter of the formulation.
2. The method for predicting an occurrence of a vehicle insurance according to claim 1, wherein after the step of inputting policy data to be predicted into the vehicle insurance occurrence prediction model to predict a predicted occurrence of whether the policy data to be predicted is an occurrence of a vehicle insurance, the method for predicting an occurrence of a vehicle insurance further comprises:
and judging whether the service time of the car insurance risk prediction model exceeds a preset period, and if so, returning to the step of acquiring a plurality of pieces of first historical policy data and corresponding historical risk results.
3. The method for predicting an emergency situation of a vehicle risk according to claim 1, wherein the first historical policy data and/or the second historical policy data includes owner information, vehicle information, insurance information.
4. The prediction system for the dangerous situation of the vehicle insurance is characterized by comprising an acquisition module, a weight calculation module, a training module and a prediction module;
the acquisition module is used for acquiring a plurality of pieces of first historical policy data and corresponding historical risk-giving results;
the weight calculation module is used for calculating the weight of each first historical policy data according to the first historical policy data and the historical risk-giving result;
the training module is used for inputting the weight and the first historical policy data into a classification model for training to obtain a car insurance risk prediction model, and output parameters of the car insurance risk prediction model comprise a predicted risk result;
the prediction module is used for inputting the policy data to be predicted into the vehicle risk and risk prediction model to predict whether the policy data to be predicted is a predicted risk and risk result;
the output parameters of the vehicle risk prediction model also comprise scoring values;
the weight calculation module is also used for calculating and obtaining the weight by adopting at least two preset calculation methods;
the training module is further used for taking the weights obtained by calculation of each preset calculation method and the first historical policy data as input to obtain respectively corresponding car insurance risk prediction models as car insurance risk prediction models to be selected;
the training module is further configured to input a plurality of second historical policy data to the to-be-selected vehicle risk prediction model respectively to obtain a corresponding prediction risk result and a scoring value, calculate a lifting degree of the vehicle risk prediction model according to the scoring value and the historical risk result corresponding to the second historical policy data, and select a model with the highest lifting degree in the to-be-selected vehicle risk prediction model as a final vehicle risk prediction model;
the promotion degree is used for representing model evaluation parameters related to the number of risk, amount or odds;
the preset calculation method comprises the following steps:
wherein weight1 is the weight, claimcnt i For the number of danger, r i For the annual proportion, α is a limiting parameter, within a preset threshold range;
and/or the number of the groups of groups,
wherein weight2 is the weight, claimcnt i For the number of danger, r i Beta is a parameter of evolution for annual proportion;
and/or the number of the groups of groups,
wherein weight3 is the weight, claimcnt i For the number of danger, r i For annual proportion, β is the parameter of the formulation.
5. The system for predicting the risk of a vehicle according to claim 4, further comprising an update module, wherein the update module is configured to determine whether a usage time of the risk prediction model exceeds a preset period, and if so, call the collection module.
6. The system for predicting an emergency situation of a vehicle risk according to claim 4, wherein the first historical policy data and/or the second historical policy data includes owner information, vehicle information, insurance information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910173365.7A CN109784586B (en) | 2019-03-07 | 2019-03-07 | Prediction method and system for danger emergence condition of vehicle danger |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910173365.7A CN109784586B (en) | 2019-03-07 | 2019-03-07 | Prediction method and system for danger emergence condition of vehicle danger |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109784586A CN109784586A (en) | 2019-05-21 |
| CN109784586B true CN109784586B (en) | 2023-08-29 |
Family
ID=66486347
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910173365.7A Active CN109784586B (en) | 2019-03-07 | 2019-03-07 | Prediction method and system for danger emergence condition of vehicle danger |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109784586B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111652614B (en) * | 2020-06-01 | 2023-08-22 | 泰康保险集团股份有限公司 | Data processing system, data processing method and device |
| CN112862621A (en) * | 2021-01-15 | 2021-05-28 | 广州亚美信息科技有限公司 | Vehicle insurance risk assessment method and device and computer equipment |
| CN113255842B (en) * | 2021-07-05 | 2021-11-02 | 平安科技（深圳）有限公司 | Vehicle replacement prediction method, device, equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106411998A (en) * | 2016-07-15 | 2017-02-15 | 南京邮电大学 | Prediction method for UBI (Usage-Based Insurance) system based on internet of vehicles big data |
| CN107292528A (en) * | 2017-06-30 | 2017-10-24 | 阿里巴巴集团控股有限公司 | Vehicle insurance Risk Forecast Method, device and server |
| CN108648011A (en) * | 2018-05-11 | 2018-10-12 | 上海赢科信息技术有限公司 | Model generates, identification client buys the method and system of vehicle insurance intention |
| CN108665175A (en) * | 2018-05-16 | 2018-10-16 | 阿里巴巴集团控股有限公司 | A kind of processing method, device and the processing equipment of insurance business risk profile |
-
2019
- 2019-03-07 CN CN201910173365.7A patent/CN109784586B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106411998A (en) * | 2016-07-15 | 2017-02-15 | 南京邮电大学 | Prediction method for UBI (Usage-Based Insurance) system based on internet of vehicles big data |
| CN107292528A (en) * | 2017-06-30 | 2017-10-24 | 阿里巴巴集团控股有限公司 | Vehicle insurance Risk Forecast Method, device and server |
| CN108648011A (en) * | 2018-05-11 | 2018-10-12 | 上海赢科信息技术有限公司 | Model generates, identification client buys the method and system of vehicle insurance intention |
| CN108665175A (en) * | 2018-05-16 | 2018-10-16 | 阿里巴巴集团控股有限公司 | A kind of processing method, device and the processing equipment of insurance business risk profile |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109784586A (en) | 2019-05-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wen et al. | Degradation modeling and RUL prediction using Wiener process subject to multiple change points and unit heterogeneity | |
| CN109784586B (en) | Prediction method and system for danger emergence condition of vehicle danger | |
| US9082072B1 (en) | Method for applying usage based data | |
| CN109670970B (en) | Driving behavior scoring method and device and computer readable storage medium | |
| CN111242484B (en) | Vehicle risk comprehensive evaluation method based on transition probability | |
| CN109658272A (en) | Driving behavior evaluation system and Insurance Pricing system based on driving behavior | |
| US20200134729A1 (en) | Information processing device, information processing system, information processing method, and program | |
| CN106649709B (en) | Vehicle stopping behavior pattern prediction and evaluation method based on data mining | |
| CN108876165B (en) | Driver safety monitoring learning system | |
| CN109543710B (en) | Model training method and related equipment | |
| CN110968839A (en) | Driving risk assessment method, device, equipment and storage medium | |
| CN113423063B (en) | Vehicle monitoring method and device based on vehicle-mounted T-BOX, vehicle and medium | |
| Meng et al. | Actuarial intelligence in auto insurance: Claim frequency modeling with driving behavior features and improved boosted trees | |
| CN115034821A (en) | Vehicle estimation method and device, computer equipment and storage medium | |
| CN114971009A (en) | Vehicle insurance risk prediction method and computer equipment | |
| CN113283548A (en) | Vehicle safety scoring method, device, equipment and storage medium | |
| CN113870020A (en) | Overdue risk control method and device | |
| CN113554517A (en) | Risk assessment method and device for truck insurance market and storage medium | |
| CN112990563A (en) | Real-time prediction method for rear-end collision accident risk of expressway | |
| CN113487109A (en) | Group identification method and device, electronic equipment and storage medium | |
| CN113807587A (en) | Integral early warning method and system based on multi-ladder-core deep neural network model | |
| CN116091254B (en) | Commercial vehicle risk analysis method | |
| Marinelli et al. | Investigation of earthmoving trucks deterioration using discriminant analysis | |
| CN116453345B (en) | Bus driving safety early warning method and system based on driving risk feedback | |
| US20240092394A1 (en) | Generating unknown-unsafe scenarios, improving automated vehicles, and computer system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |