CN116777642A - Vehicle risk parameter prediction method and device based on ensemble learning model - Google Patents
Vehicle risk parameter prediction method and device based on ensemble learning model Download PDFInfo
- Publication number
- CN116777642A CN116777642A CN202310741089.6A CN202310741089A CN116777642A CN 116777642 A CN116777642 A CN 116777642A CN 202310741089 A CN202310741089 A CN 202310741089A CN 116777642 A CN116777642 A CN 116777642A
- Authority
- CN
- China
- Prior art keywords
- learning model
- risk
- vehicle
- basic
- prediction
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 72
- 238000012795 verification Methods 0.000 claims abstract description 125
- 238000012549 training Methods 0.000 claims abstract description 82
- 238000012216 screening Methods 0.000 claims abstract description 17
- 238000012545 processing Methods 0.000 claims description 17
- 238000010801 machine learning Methods 0.000 claims description 16
- 230000006399 behavior Effects 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 5
- 230000000694 effects Effects 0.000 abstract description 8
- 238000005516 engineering process Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 7
- 238000013473 artificial intelligence Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 238000002790 cross-validation Methods 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000010200 validation analysis Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 235000019800 disodium phosphate Nutrition 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/08—Insurance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
Abstract
The embodiment of the application belongs to the technical field of intelligent decision making, and relates to a vehicle risk parameter prediction method and device based on an integrated learning model, wherein the method comprises the steps of inputting a prediction sample set into a trained integrated learning model to predict the vehicle risk parameter, and obtaining a prediction result; and acquiring the risk parameters of the risk of each target vehicle in the prediction sample set according to the vehicle distinguishing identification information and the prediction result. The basic learning model is trained by using cross verification, and due to the random screening property of the training sample subset, the basic learning model is obtained by training according to different training sample sets and basic verification sets, so that the training is more scientific, and meanwhile, the weight vector of each basic learning model is adjusted by carrying out iterative verification on the super learning model, so that the influence of a certain basic learning model on the output effect of the final integrated learning model due to insufficient precision or over fitting problem is avoided to the greatest extent, and the scientificity and the accuracy of the risk parameter prediction result of the vehicle risk are further ensured.
Description
Technical Field
The application relates to the technical field of intelligent decision making, in particular to a vehicle risk parameter prediction method based on an integrated learning model and related equipment thereof.
Background
There are two methods for predicting risk parameters of a vehicle risk: one is to calculate the prediction of the pure risk premium by multiplying the prediction of the risk frequency and the prediction of the case average claim under the assumption that the risk frequency and the case average claim are independent. The disadvantage of this approach is that the possible dependency between the frequency of the risk of emergence and the case-average claim is ignored. Thus, a generalized linear model is used to predict the risk parameters for the vehicle risk.
Currently, due to the continuous change of market environment, a higher requirement is placed on the prediction accuracy of the model, and only when an insurance company has a more accurate estimation of the pay cost behind a policy relative to a competitor, the corresponding underwriting risk can be compensated by reasonable premium pricing and service investment. Because the modeling features have multiple types, only a single generalized linear model is adopted to predict the risk parameters of the vehicle risk, and certain limitation exists, so that the accuracy of a prediction result cannot be ensured. Therefore, the prior art also has a certain limitation in predicting the risk parameters of the vehicle risk, and the accuracy of the prediction result cannot be ensured.
Disclosure of Invention
The embodiment of the application aims to provide a vehicle risk parameter prediction method based on an integrated learning model and related equipment thereof, so as to solve the problems that the prediction of the vehicle risk parameter in the prior art has certain limitation and the accuracy of a prediction result cannot be ensured.
In order to solve the technical problems, the embodiment of the application provides a vehicle risk parameter prediction method based on an ensemble learning model, which adopts the following technical scheme:
a vehicle risk parameter prediction method based on an ensemble learning model comprises the following steps:
obtaining a prediction sample set, wherein the prediction sample set is a sample set without marked risk parameter information of the vehicle risk, and each sample in the prediction sample set contains influence factor information for predicting the risk parameter of the vehicle risk, wherein the influence factor information comprises current value information, service life information, historical risk record information, region range information for the vehicle to travel frequently and driving behavior information of a driver of a target vehicle;
inputting the prediction sample set into a trained integrated learning model to predict risk parameters of the vehicle risk, and obtaining a prediction result;
And acquiring risk parameters of the vehicle risk of each target vehicle in the prediction sample set according to the vehicle distinguishing identification information and the prediction result, wherein the vehicle distinguishing identification information comprises license plate number or frame number information.
Further, before performing the step of obtaining a set of prediction samples, the method further comprises:
acquiring a sample set of marked vehicle risk parameter information, wherein each sample in the sample set of marked vehicle risk parameter information contains actual vehicle risk parameter information and influence factor information, and the influence factor information comprises current value information, service life information, historical risk record information, region range information where the vehicle frequently runs and driving behavior information of a driver of a target vehicle;
dividing the sample set of the marked risk parameter information into a training sample set and a verification sample set according to a preset proportion.
Further, before the step of inputting the prediction sample set into a trained ensemble learning model to predict risk parameters of the vehicle risk and obtain a prediction result, the method further includes:
acquiring an initial weight vector set for each trained basic learning model;
According to the initial weight vector set for each trained basic learning model, weighting and combining each trained basic learning model by adopting a model stacking mode so as to generate a super learning model;
inputting the verification sample set into the super learning model for iterative verification, and obtaining an output result of the super learning model during each iterative verification;
acquiring the loss degree of the corresponding output result in each iteration verification relative to the risk parameter information of the vehicle risk contained in each sample in the verification sample set through a preset loss function;
until iteration verification is finished, selecting a weight vector corresponding to each basic learning model when the loss degree is the minimum value as the weight vector of the corresponding basic learning model in the super learning model, and finishing tuning of the super learning model, wherein the condition for finishing the iteration verification comprises reaching a preset maximum iteration number or the loss degree is smaller than a preset expected loss value;
and setting the super learning model with the optimized tuning as the trained integrated learning model.
Further, before performing the step of obtaining the initial weight vector set for each trained base learning model, the method further includes:
Different machine learning algorithms are adopted to pre-construct a plurality of basic learning models;
the k-fold cross verification mode is adopted, the training sample set is sequentially input into the plurality of basic learning models, basic learning model training is carried out, and each trained basic learning model is obtained, and the specific implementation mode is as follows:
step 401, dividing the training sample set into N training sample subsets in an equal ratio, wherein N is a positive integer;
step 402, randomly screening any training sample subset to be used as a basic verification set;
step 403, inputting other non-screened training sample subsets into a current basic learning model, and training the current basic learning model according to a built-in machine learning algorithm in the current basic learning model to obtain a pre-trained basic learning model;
step 404, inputting the basic verification set into the pre-trained basic learning model, and performing verification tuning processing on the pre-trained basic learning model to obtain a tuned basic learning model;
step 405, using the tuned basic learning model as a currently trained basic learning model;
step 406, sequentially screening a basic learning model from the pre-constructed plurality of basic learning models as the current basic learning model, and circularly executing steps 402 to 405 to obtain each trained basic learning model.
Further, the step of inputting the basic verification set into the pre-trained basic learning model, performing verification tuning processing on the pre-trained basic learning model, and obtaining a tuned basic learning model specifically includes:
step 501, inputting the basic verification set into the pre-trained basic learning model, and obtaining risk parameter information of the vehicle risk corresponding to the basic verification set output by the pre-trained basic learning model according to the influence factor information in the basic verification set;
step 502, comparing the risk parameter information of the vehicle risk corresponding to the basic verification set output by the pre-trained basic learning model with the risk parameter information of the marked vehicle risk corresponding to the basic verification set, and obtaining a similarity comparison result;
step 503, judging whether the similarity comparison result meets a preset similarity threshold;
step 504, if the similarity comparison result does not meet the preset similarity threshold, adjusting the weight value of each influence factor in the pre-trained basic learning model, and re-executing steps 501 to 503 after adjustment;
and step 505, completing verification tuning processing of the pre-trained basic learning model until the similarity comparison result meets a preset similarity threshold value, and obtaining a tuned basic learning model.
Further, the step of inputting the verification sample set into the super learning model to perform iterative verification to obtain an output result of the super learning model during each iterative verification specifically includes:
inputting the verification sample set into the super learning model, and obtaining risk parameter information of the vehicle risk output by the super learning model according to the influence factor information corresponding to each sample in the verification sample set;
and taking the risk parameter information of the vehicle risk output by the super learning model as an output result of the super learning model when iterative verification is carried out at the current time.
Further, the step of inputting the prediction sample set into a trained ensemble learning model to predict risk parameters of the vehicle risk and obtain a prediction result specifically includes:
inputting the prediction sample set into a trained integrated learning model;
the integrated learning model carries out model prediction according to the influence factor information corresponding to each sample in the prediction sample set to obtain the risk parameter information of the vehicle risk corresponding to each sample output by the integrated learning model;
and taking the risk parameter information of the car risk corresponding to each sample output by the integrated learning model as the prediction result.
In order to solve the technical problems, the embodiment of the application also provides a vehicle risk parameter prediction device based on an ensemble learning model, which adopts the following technical scheme:
an integrated learning model-based vehicle risk parameter prediction device, comprising:
the prediction sample set acquisition module is used for acquiring a prediction sample set, wherein the prediction sample set is a sample set without marked risk parameter information of the vehicle risk, and each sample in the prediction sample set contains influence factor information for predicting the risk parameter of the vehicle risk, wherein the influence factor information comprises current value information, service life information, history risk record information, region range information for the vehicle to travel frequently and driving behavior information of a driver of a target vehicle;
the integrated learning model prediction module is used for inputting the prediction sample set into a trained integrated learning model to predict risk parameters of the vehicle to obtain a prediction result;
the prediction result acquisition module is used for acquiring the risk parameters of the vehicle risk of each target vehicle in the prediction sample set according to the vehicle distinguishing identification information and the prediction result, wherein the vehicle distinguishing identification information comprises license plate numbers or frame number information.
In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical schemes:
a computer device comprising a memory and a processor, wherein computer readable instructions are stored in the memory, and the processor implements the steps of the integrated learning model-based risk parameter prediction method described above when executing the computer readable instructions.
In order to solve the above technical problems, an embodiment of the present application further provides a computer readable storage medium, which adopts the following technical schemes:
a computer readable storage medium having stored thereon computer readable instructions which when executed by a processor implement the steps of the ensemble learning model based risk parameter prediction method as described above.
Compared with the prior art, the embodiment of the application has the following main beneficial effects:
according to the vehicle risk parameter prediction method based on the integrated learning model, a prediction sample set is obtained; inputting the prediction sample set into a trained integrated learning model to predict risk parameters of the vehicle risk, and obtaining a prediction result; and acquiring the risk parameters of the risk of each target vehicle in the prediction sample set according to the vehicle distinguishing identification information and the prediction result. Compared with a regression model for predicting pure risk premium, the method has the advantages that the basic learning model is trained by using cross verification, the influence of a division method on a training data set can be reduced as far as possible due to the random screening property of a training sample subset, and the condition that each basic learning model adopts the same training sample set and basic verification set is avoided, so that the basic learning model is obtained by training according to different training sample sets and basic verification sets, the method is more scientific, and meanwhile, the weight vector of each basic learning model is adjusted by carrying out iterative verification on a super learning model, so that the influence of insufficient precision or overfitting problem of a certain basic learning model on the output effect of a final integrated learning model is avoided to the greatest extent, and the scientificity and the accuracy of a risk parameter prediction result are further ensured.
Drawings
In order to more clearly illustrate the solution of the present application, a brief description will be given below of the drawings required for the description of the embodiments of the present application, it being apparent that the drawings in the following description are some embodiments of the present application, and that other drawings may be obtained from these drawings without the exercise of inventive effort for a person of ordinary skill in the art.
FIG. 1 is an exemplary system architecture diagram in which the present application may be applied;
FIG. 2 is a flow chart of one embodiment of a vehicle risk parameter prediction method based on an ensemble learning model in accordance with the present application;
FIG. 3 is a flow chart of one embodiment of obtaining a trained ensemble learning model in a ensemble learning model-based risk parameter prediction method in accordance with the present application;
FIG. 4 is a flow chart of one embodiment of obtaining a trained base learning model in an ensemble learning model-based risk parameter prediction method in accordance with the present application;
FIG. 5 is a flow chart of one embodiment of step 404 shown in FIG. 4;
FIG. 6 is a flow chart of one embodiment of step 303 shown in FIG. 3;
FIG. 7 is a flow chart of one embodiment of step 202 of FIG. 2;
FIG. 8 is a schematic structural view of one embodiment of an ensemble learning model-based risk parameter prediction apparatus according to the present application;
FIG. 9 is a schematic structural diagram of a specific embodiment of the ensemble learning model generation module in an ensemble learning model-based risk parameter prediction apparatus according to the present application;
FIG. 10 is a schematic diagram of one embodiment of the k-fold cross-validation training sub-module in an ensemble learning model based risk parameter prediction device in accordance with the present application;
FIG. 11 is a schematic diagram of an embodiment of a computer device in accordance with the application.
Detailed Description
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used in the description of the applications herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having" and any variations thereof in the description of the application and the claims and the description of the drawings above are intended to cover a non-exclusive inclusion. The terms first, second and the like in the description and in the claims or in the above-described figures, are used for distinguishing between different objects and not necessarily for describing a sequential or chronological order.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.
In order to make the person skilled in the art better understand the solution of the present application, the technical solution of the embodiment of the present application will be clearly and completely described below with reference to the accompanying drawings.
As shown in fig. 1, a system architecture 100 may include terminal devices 101, 102, 103, a network 104, and a server 105. The network 104 is used as a medium to provide communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, among others.
The user may interact with the server 105 via the network 104 using the terminal devices 101, 102, 103 to receive or send messages or the like. Various communication client applications, such as a web browser application, a shopping class application, a search class application, an instant messaging tool, a mailbox client, social platform software, etc., may be installed on the terminal devices 101, 102, 103.
The terminal devices 101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smartphones, tablet computers, electronic book readers, MP3 players (Moving Picture ExpertsGroup Audio Layer III, dynamic video expert compression standard audio plane 3), MP4 (Moving PictureExperts Group Audio Layer IV, dynamic video expert compression standard audio plane 4) players, laptop and desktop computers, and the like.
The server 105 may be a server providing various services, such as a background server providing support for pages displayed on the terminal devices 101, 102, 103.
It should be noted that, the vehicle risk parameter prediction method based on the integrated learning model provided by the embodiment of the application is generally executed by a server/terminal device, and correspondingly, the vehicle risk parameter prediction device based on the integrated learning model is generally arranged in the server/terminal device.
It should be understood that the number of terminal devices, networks and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
With continued reference to fig. 2, a flow chart of one embodiment of a vehicle risk parameter prediction method based on an ensemble learning model in accordance with the present application is shown. The vehicle risk parameter prediction method based on the integrated learning model comprises the following steps:
In step 201, a set of prediction samples is obtained.
In this embodiment, the prediction sample set is a sample set not marked with risk parameter information of the vehicle risk, where the risk parameter of the vehicle risk specifically refers to a risk premium, and the risk premium of the vehicle risk is also referred to as "pure premium" of the vehicle risk and refers to an expected premium that completely meets the requirements of claim expenditure within a guarantee period.
In this embodiment, each sample in the prediction sample set includes an influence factor information for predicting a risk parameter of a vehicle, where the influence factor information includes current value information of a target vehicle, age information of the target vehicle, history information of a risk record, region range information of a frequent running of the vehicle, and driving behavior information of a driver.
In this embodiment, before performing the step of obtaining the prediction sample set, the method further includes: acquiring a sample set of marked risk parameter information of the vehicle risk; dividing the sample set of the marked risk parameter information into a training sample set and a verification sample set according to a preset proportion.
In this embodiment, each sample in the sample set of marked risk parameter information includes an actual risk parameter information and the influence factor information, where the influence factor information includes current value information, used age information, historical risk record information, region range information where the vehicle is often driven, and driving behavior information of the driver.
When predicting the risk parameters of the vehicle insurance, the current value information of the vehicle can influence the selection of the insurance amount, the insurance amount of the vehicle damage and the theft and rescue should be equal to the value of the vehicle, and the value of the vehicle is higher than that of the vehicle and is invalid and lower than that of the vehicle and cannot be effectively ensured. The age information of the vehicle also affects the risk of driving the vehicle and the current value information of the vehicle, thereby affecting the risk parameters of the vehicle risk. Furthermore, factors affecting driving safety and increasing traffic accident risk can all affect risk parameters of the vehicle risk, such as driving level and safety awareness of a driver of the vehicle, historical risk record information, regional range information of frequent driving of the vehicle, driving behavior information of the driver and the like. Therefore, the embodiment of the application takes the current value information, the service life information, the historical risk record information, the regional range information of the frequent running of the vehicle and the driving behavior information of the driver of the target vehicle as the influence factor information of the prediction of the risk parameter information of the vehicle risk preferentially. In the prediction of risk parameters of the car insurance, the value information of the car insurance itself is considered in the selection of influencing factors, the driving behavior information of the car owner is also considered, and the scientificity and the accuracy of the prediction are ensured.
Step 202, inputting the prediction sample set into a trained integrated learning model to predict risk parameters of the vehicle risk, and obtaining a prediction result.
In this embodiment, before the step of inputting the prediction sample set into a trained integrated learning model to perform risk parameter prediction, and obtaining a prediction result, the method further includes obtaining the trained integrated learning model.
With continued reference to fig. 3, fig. 3 is a flowchart of an embodiment of acquiring a trained ensemble learning model in an ensemble learning model-based risk parameter prediction method according to the present application, including:
step 301, obtaining an initial weight vector set for each trained basic learning model;
in this embodiment, before performing the step of obtaining the initial weight vector set for each trained basic learning model, the method further includes: different machine learning algorithms are adopted to pre-construct a plurality of basic learning models; and sequentially inputting the training sample set into the plurality of basic learning models by adopting a k-fold cross verification mode, and training the basic learning models to obtain each trained basic learning model.
By adopting different machine learning algorithms, a plurality of basic learning models are pre-built, the fact that the machine learning algorithms used by the machine learning models have certain distinguishability due to the fact that the influence factor information is different is fully considered, the difference between the influence factor information is fully considered is reflected, the basic learning models are built by using different machine learning algorithms, the fact that all the basic learning models adopt the same machine learning algorithm is avoided, the fact that the output result of the final integrated learning model is obtained by comprehensively operating different machine learning algorithms is guaranteed, and the machine learning model is more scientific.
With continued reference to the drawings, fig. 4 is a flowchart of one embodiment of acquiring a trained base learning model in an integrated learning model-based risk parameter prediction method according to the present application, including:
step 401, dividing the training sample set into N training sample subsets in an equal ratio, wherein N is a positive integer;
step 402, randomly screening any training sample subset to be used as a basic verification set;
step 403, inputting other non-screened training sample subsets into a current basic learning model, and training the current basic learning model according to a built-in machine learning algorithm in the current basic learning model to obtain a pre-trained basic learning model;
The basic learning model is trained by using cross verification, and due to the random screening property of the training sample subsets, the influence of the division method on the training data set can be reduced as far as possible, and the condition that each basic learning model adopts the same training sample set and basic verification set is avoided, so that the basic learning model is obtained by training according to different training sample sets and basic verification sets, and the basic learning model is more scientific.
Step 404, inputting the basic verification set into the pre-trained basic learning model, and performing verification tuning processing on the pre-trained basic learning model to obtain a tuned basic learning model;
with continued reference to fig. 5, fig. 5 is a flow chart of one embodiment of step 404 shown in fig. 4, comprising:
step 501, inputting the basic verification set into the pre-trained basic learning model, and obtaining risk parameter information of the vehicle risk corresponding to the basic verification set output by the pre-trained basic learning model according to the influence factor information in the basic verification set;
step 502, comparing the risk parameter information of the vehicle risk corresponding to the basic verification set output by the pre-trained basic learning model with the risk parameter information of the marked vehicle risk corresponding to the basic verification set, and obtaining a similarity comparison result;
Step 503, judging whether the similarity comparison result meets a preset similarity threshold;
step 504, if the similarity comparison result does not meet the preset similarity threshold, adjusting the weight value of each influence factor in the pre-trained basic learning model, and re-executing steps 501 to 503 after adjustment;
and step 505, completing verification tuning processing of the pre-trained basic learning model until the similarity comparison result meets a preset similarity threshold value, and obtaining a tuned basic learning model.
And performing verification and optimization processing on the pre-trained basic learning model through the basic verification set to obtain an optimized basic learning model, wherein the basic learning model is substantially obtained by performing similarity comparison with the risk parameter information of the risk marked in the basic verification set and the risk parameter information of the risk marked in the basic verification set, and when the similarity does not meet the preset acquaintance threshold requirement, the output result of the basic learning model finally obtained meets the preset acquaintance threshold requirement by adjusting the weight value of each influence factor in the basic learning model. By optimizing the basic learning model, the scientificity and the accuracy of the prediction result of the risk parameters of the vehicle risk are ensured.
Step 405, using the tuned basic learning model as a currently trained basic learning model;
step 406, sequentially screening a basic learning model from the pre-constructed plurality of basic learning models as the current basic learning model, and circularly executing steps 402 to 405 to obtain each trained basic learning model.
And (3) through circularly executing the steps 402 to 405, each trained basic learning model is obtained, so that the basic verification set is randomly screened once again each time when one basic learning model is screened out as the current basic learning model, and the difference between the training sample set used by the basic learning model and the basic verification set is ensured as much as possible.
Step 302, weighting and combining each trained basic learning model by adopting a model stacking mode according to the initial weight vector set for each trained basic learning model so as to generate a super learning model;
step 303, inputting the verification sample set into the super learning model, and performing iterative verification to obtain an output result of the super learning model during each iterative verification;
with continued reference to fig. 6, fig. 6 is a flow chart of one embodiment of step 303 shown in fig. 3, comprising:
Step 601, inputting the verification sample set into the super learning model, and obtaining risk parameter information of the vehicle risk output by the super learning model according to the influence factor information corresponding to each sample in the verification sample set;
and 602, taking the risk parameter information of the vehicle risk output by the super learning model as an output result of the super learning model when iterative verification is performed at the current time.
Step 304, obtaining the loss degree of the corresponding output result in each iteration verification relative to the risk parameter information of the vehicle risk contained in each sample in the verification sample set through a preset loss function;
step 305, until the iteration verification is finished, selecting a weight vector corresponding to each basic learning model when the loss degree is the minimum as a weight vector of a corresponding basic learning model in the super learning model, and finishing tuning of the super learning model, wherein the condition of finishing the iteration verification comprises reaching a preset maximum iteration number or the loss degree is smaller than a preset expected loss value;
and 306, acquiring the super learning model with the optimized function and setting the super learning model as the trained integrated learning model.
And (3) through iterative verification on the super learning model, adjusting the weight vector of each basic learning model, and maximally avoiding the influence of insufficient precision or over-fitting problem of a certain basic learning model on the output effect of the final integrated learning model.
With continued reference to fig. 7, fig. 7 is a flow chart of one embodiment of step 202 of fig. 2, including:
step 701, inputting the prediction sample set into a trained integrated learning model;
step 702, the integrated learning model performs model prediction according to the influence factor information corresponding to each sample in the prediction sample set, so as to obtain risk parameter information of vehicle risk corresponding to each sample output by the integrated learning model;
and step 703, taking the risk parameter information of the car risk corresponding to each sample output by the integrated learning model as the prediction result.
And 203, acquiring risk parameters of the vehicle risk of each target vehicle in the prediction sample set according to the vehicle distinguishing identification information and the prediction result.
In this embodiment, the vehicle distinguishing identification information includes license plate number or frame number information.
And associating the risk parameters of the vehicle risk corresponding to each sample in the prediction sample set with the corresponding vehicle through the vehicle distinguishing identification information.
The method comprises the steps of obtaining a prediction sample set; inputting the prediction sample set into a trained integrated learning model to predict risk parameters of the vehicle risk, and obtaining a prediction result; and acquiring the risk parameters of the risk of each target vehicle in the prediction sample set according to the vehicle distinguishing identification information and the prediction result. Compared with a regression model for predicting pure risk premium, the method has the advantages that the basic learning model is trained by using cross verification, the influence of a division method on a training data set can be reduced as far as possible due to the random screening property of a training sample subset, and the condition that each basic learning model adopts the same training sample set and basic verification set is avoided, so that the basic learning model is obtained by training according to different training sample sets and basic verification sets, the method is more scientific, and meanwhile, the weight vector of each basic learning model is adjusted by carrying out iterative verification on a super learning model, so that the influence of insufficient precision or overfitting problem of a certain basic learning model on the output effect of a final integrated learning model is avoided to the greatest extent, and the scientificity and the accuracy of a risk parameter prediction result are further ensured.
The embodiment of the application can acquire and process the related data based on the artificial intelligence technology. Among these, artificial intelligence (Artificial Intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a digital computer-controlled machine to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use knowledge to obtain optimal results.
Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and other directions.
In the embodiment of the application, the basic learning model is trained by using cross verification, and the influence of a division method on a training data set can be reduced as far as possible due to the random screening property of a training sample subset, so that each basic learning model adopts the same training sample set and basic verification set, the basic learning model is obtained by training according to different training sample sets and basic verification sets, the basic learning model is more scientific, and simultaneously, the weight vector of each basic learning model is adjusted by carrying out iterative verification on the super learning model, the influence of insufficient precision or over-fitting problem of a certain basic learning model on the output effect of the final integrated learning model is avoided to the greatest extent, and the scientificity and the accuracy of the risk parameter prediction result are further ensured.
With further reference to fig. 8, as an implementation of the method shown in fig. 2, the present application provides an embodiment of a vehicle risk parameter prediction apparatus based on an ensemble learning model, where the embodiment of the apparatus corresponds to the embodiment of the method shown in fig. 2, and the apparatus may be specifically applied to various electronic devices.
As shown in fig. 8, the vehicle risk parameter prediction apparatus 800 based on the ensemble learning model according to the present embodiment includes: a prediction sample set acquisition module 801, an ensemble learning model prediction module 802, and a prediction result acquisition module 803. Wherein:
a prediction sample set obtaining module 801, configured to obtain a prediction sample set, where the prediction sample set is a sample set not labeled with risk parameter information of a vehicle risk, and each sample in the prediction sample set includes influence factor information for predicting the risk parameter of the vehicle risk, where the influence factor information includes current value information, age information, history information of a risk record, region range information where the vehicle is often driven, and driving behavior information of a driver;
the integrated learning model prediction module 802 is configured to input the prediction sample set into a trained integrated learning model to predict risk parameters of a vehicle, so as to obtain a prediction result;
The prediction result obtaining module 803 is configured to obtain risk parameters of each target vehicle in the prediction sample set according to vehicle identification information and the prediction result, where the vehicle identification information includes license plate number or frame number information.
In some embodiments of the present application, the vehicle risk parameter prediction apparatus 800 based on the ensemble learning model further includes a labeled sample acquiring module and a labeled sample processing module.
Wherein:
the marked sample acquisition module is used for acquiring a sample set of marked risk parameter information of the vehicle risk;
the marked sample processing module is used for dividing the sample set of the marked risk parameter information of the vehicle risk into a training sample set and a verification sample set according to a preset proportion.
With continued reference to fig. 9, in some embodiments of the present application, the ensemble learning model-based risk parameter prediction apparatus 800 further includes an ensemble learning model generating module 804, and fig. 9 is a schematic structural diagram of a specific embodiment of the ensemble learning model generating module 804 in the ensemble learning model-based risk parameter prediction apparatus according to the present application, as shown in fig. 9, where the ensemble learning model generating module 804 includes: an initial weight vector acquisition sub-module 901, a basic learning model stacking sub-module 902, an iterative verification sub-module 903, a loss degree acquisition sub-module 904, a weight vector preference sub-module 905, and an integrated learning model final acquisition sub-module 906. Wherein:
An initial weight vector obtaining sub-module 901, configured to obtain an initial weight vector set for each trained basic learning model;
a basic learning model stacking sub-module 902, configured to weight-combine each trained basic learning model in a model stacking manner according to the initial weight vector set for each trained basic learning model, so as to generate a super learning model;
the iteration verification sub-module 903 is configured to input the verification sample set into the super learning model, perform iteration verification, and obtain an output result of the super learning model during each iteration verification;
the loss degree obtaining sub-module 904 is configured to obtain, according to a preset loss function, a loss degree of the corresponding output result in each iteration verification with respect to the risk parameter information of the vehicle risk contained in each sample in the verification sample set;
the weight vector optimization sub-module 905 is configured to select, until the iteration verification is finished, a weight vector corresponding to each basic learning model when the loss degree is the minimum value as a weight vector of a corresponding basic learning model in the super learning model, and complete tuning of the super learning model, where a condition that the iteration verification is finished includes reaching a preset maximum iteration number or that the loss degree is less than a preset expected loss value;
And the integrated learning model final acquisition sub-module 906 is configured to acquire the super learning model with the tuned optimization being set as the trained integrated learning model.
In some embodiments of the present application, the integrated learning model-based risk parameter prediction apparatus 800 further includes a basic learning model training module, where the basic learning model training module includes a basic learning model pre-building sub-module and a k-fold cross-validation training sub-module.
Wherein:
the basic learning model pre-constructing sub-module is used for pre-constructing a plurality of basic learning models by adopting different machine learning algorithms;
with continued reference to fig. 10, fig. 10 is a schematic structural diagram of an example of the k-fold cross-validation training sub-module in the integrated learning model-based risk parameter prediction apparatus according to the present application, and as shown in fig. 10, in some embodiments of the present application, the k-fold cross-validation training sub-module further includes an equal ratio dividing unit 10a, a basic validation set random screening unit 10b, a basic learning model training unit 10c, a basic learning model validation tuning unit 10d, a basic learning model training completion unit 10e, and a basic learning model circulation training unit 10f. Wherein:
An equal ratio dividing unit 10a, configured to divide the training sample set into N training sample subsets in equal ratio, where N is a positive integer;
a basic verification set random screening unit 10b, configured to randomly screen any one of the training sample subsets as a basic verification set;
a basic learning model training unit 10c, configured to input a subset of training samples that are not screened into a current basic learning model, and train the current basic learning model according to a built-in machine learning algorithm in the current basic learning model to obtain a pre-trained basic learning model;
a basic learning model verification tuning unit 10d, configured to input the basic verification set into the pre-trained basic learning model, perform verification tuning processing on the pre-trained basic learning model, and obtain a tuned basic learning model;
a basic learning model training completion unit 10e, configured to use the tuned basic learning model as a currently trained basic learning model;
and the basic learning model circulation training unit 10f is used for sequentially screening one basic learning model from the plurality of pre-constructed basic learning models to be used as the current basic learning model, and training the current basic learning model to obtain each trained basic learning model.
The method comprises the steps of obtaining a prediction sample set; inputting the prediction sample set into a trained integrated learning model to predict risk parameters of the vehicle risk, and obtaining a prediction result; and acquiring the risk parameters of the risk of each target vehicle in the prediction sample set according to the vehicle distinguishing identification information and the prediction result. Compared with a regression model for predicting pure risk premium, the method has the advantages that the basic learning model is trained by using cross verification, the influence of a division method on a training data set can be reduced as far as possible due to the random screening property of a training sample subset, and the condition that each basic learning model adopts the same training sample set and basic verification set is avoided, so that the basic learning model is obtained by training according to different training sample sets and basic verification sets, the method is more scientific, and meanwhile, the weight vector of each basic learning model is adjusted by carrying out iterative verification on a super learning model, so that the influence of insufficient precision or overfitting problem of a certain basic learning model on the output effect of a final integrated learning model is avoided to the greatest extent, and the scientificity and the accuracy of a risk parameter prediction result are further ensured.
Those skilled in the art will appreciate that implementing all or part of the above described embodiment methods may be accomplished by computer readable instructions, stored on a computer readable storage medium, that the program when executed may comprise the steps of embodiments of the methods described above. The storage medium may be a nonvolatile storage medium such as a magnetic disk, an optical disk, a Read-Only Memory (ROM), or a random access Memory (Random Access Memory, RAM).
It should be understood that, although the steps in the flowcharts of the figures are shown in order as indicated by the arrows, these steps are not necessarily performed in order as indicated by the arrows. The steps are not strictly limited in order and may be performed in other orders, unless explicitly stated herein. Moreover, at least some of the steps in the flowcharts of the figures may include a plurality of sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, the order of their execution not necessarily being sequential, but may be performed in turn or alternately with other steps or at least a portion of the other steps or stages.
In order to solve the technical problems, the embodiment of the application also provides computer equipment. Referring specifically to fig. 11, fig. 11 is a basic structural block diagram of a computer device according to the present embodiment.
The computer device 11 comprises a memory 11a, a processor 11b, a network interface 11c communicatively connected to each other via a system bus. It should be noted that only computer device 11 having components 11a-11c is shown in the figures, but it should be understood that not all of the illustrated components need be implemented and that more or fewer components may alternatively be implemented. It will be appreciated by those skilled in the art that the computer device herein is a device capable of automatically performing numerical calculations and/or information processing in accordance with predetermined or stored instructions, the hardware of which includes, but is not limited to, microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASICs), programmable gate arrays (fields-Programmable Gate Array, FPGAs), digital processors (Digital Signal Processor, DSPs), embedded devices, etc.
The computer equipment can be a desktop computer, a notebook computer, a palm computer, a cloud server and other computing equipment. The computer equipment can perform man-machine interaction with a user through a keyboard, a mouse, a remote controller, a touch pad or voice control equipment and the like.
The memory 11a includes at least one type of readable storage medium including flash memory, hard disk, multimedia card, card memory (e.g., SD or DX memory, etc.), random Access Memory (RAM), static Random Access Memory (SRAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), programmable Read Only Memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the storage 11a may be an internal storage unit of the computer device 11, such as a hard disk or a memory of the computer device 11. In other embodiments, the memory 11a may also be an external storage device of the computer device 11, for example, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash Card (Flash Card) or the like, which are provided on the computer device 11. Of course, the memory 11a may also include both an internal memory unit of the computer device 11 and an external memory device. In this embodiment, the memory 11a is generally used for storing an operating system and various application software installed on the computer device 11, such as computer readable instructions of a vehicle risk parameter prediction method based on an ensemble learning model. Further, the memory 11a may be used to temporarily store various types of data that have been output or are to be output.
The processor 11b may be a central processing unit (Central Processing Unit, CPU), controller, microcontroller, microprocessor, or other data processing chip in some embodiments. The processor 11b is typically used to control the overall operation of the computer device 11. In this embodiment, the processor 11b is configured to execute computer readable instructions stored in the memory 11a or process data, such as computer readable instructions for executing the integrated learning model-based risk parameter prediction method.
The network interface 11c may comprise a wireless network interface or a wired network interface, which network interface 11c is typically used to establish a communication connection between the computer device 11 and other electronic devices.
The embodiment provides computer equipment, which belongs to the technical field of intelligent decision making. The method comprises the steps of obtaining a prediction sample set; inputting the prediction sample set into a trained integrated learning model to predict risk parameters of the vehicle risk, and obtaining a prediction result; and acquiring the risk parameters of the risk of each target vehicle in the prediction sample set according to the vehicle distinguishing identification information and the prediction result. Compared with a regression model for predicting pure risk premium, the method has the advantages that the basic learning model is trained by using cross verification, the influence of a division method on a training data set can be reduced as far as possible due to the random screening property of a training sample subset, and the condition that each basic learning model adopts the same training sample set and basic verification set is avoided, so that the basic learning model is obtained by training according to different training sample sets and basic verification sets, the method is more scientific, and meanwhile, the weight vector of each basic learning model is adjusted by carrying out iterative verification on a super learning model, so that the influence of insufficient precision or overfitting problem of a certain basic learning model on the output effect of a final integrated learning model is avoided to the greatest extent, and the scientificity and the accuracy of a risk parameter prediction result are further ensured.
The present application also provides another embodiment, namely, a computer readable storage medium, where computer readable instructions are stored, where the computer readable instructions are executable by a processor, so that the processor performs the steps of the vehicle risk parameter prediction method based on the integrated learning model.
The embodiment provides a computer readable storage medium, which belongs to the technical field of intelligent decision making. The method comprises the steps of obtaining a prediction sample set; inputting the prediction sample set into a trained integrated learning model to predict risk parameters of the vehicle risk, and obtaining a prediction result; and acquiring the risk parameters of the risk of each target vehicle in the prediction sample set according to the vehicle distinguishing identification information and the prediction result. Compared with a regression model for predicting pure risk premium, the method has the advantages that the basic learning model is trained by using cross verification, the influence of a division method on a training data set can be reduced as far as possible due to the random screening property of a training sample subset, and the condition that each basic learning model adopts the same training sample set and basic verification set is avoided, so that the basic learning model is obtained by training according to different training sample sets and basic verification sets, the method is more scientific, and meanwhile, the weight vector of each basic learning model is adjusted by carrying out iterative verification on a super learning model, so that the influence of insufficient precision or overfitting problem of a certain basic learning model on the output effect of a final integrated learning model is avoided to the greatest extent, and the scientificity and the accuracy of a risk parameter prediction result are further ensured.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) comprising instructions for causing a terminal device (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the method according to the embodiments of the present application.
It is apparent that the above-described embodiments are only some embodiments of the present application, but not all embodiments, and the preferred embodiments of the present application are shown in the drawings, which do not limit the scope of the patent claims. This application may be embodied in many different forms, but rather, embodiments are provided in order to provide a thorough and complete understanding of the present disclosure. Although the application has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments described in the foregoing description, or equivalents may be substituted for elements thereof. All equivalent structures made by the content of the specification and the drawings of the application are directly or indirectly applied to other related technical fields, and are also within the scope of the application.
Claims (10)
1. The vehicle risk parameter prediction method based on the ensemble learning model is characterized by comprising the following steps of:
obtaining a prediction sample set, wherein the prediction sample set is a sample set without marked risk parameter information of the vehicle risk, each sample in the prediction sample set contains influence factor information for predicting the risk parameter of the vehicle risk, and the influence factor information comprises current value information, service life information, historical risk record information, region range information for the vehicle to frequently run and driving behavior information of a driver of a target vehicle;
inputting the prediction sample set into a trained integrated learning model to predict risk parameters of the vehicle risk, and obtaining a prediction result;
and acquiring risk parameters of the vehicle risk of each target vehicle in the prediction sample set according to the vehicle distinguishing identification information and the prediction result, wherein the vehicle distinguishing identification information comprises license plate number or frame number information.
2. The ensemble learning model-based risk parameter prediction method of vehicle risk, as set forth in claim 1, characterized in that before executing the step of obtaining a prediction sample set, the method further includes:
acquiring a sample set of marked vehicle risk parameter information, wherein each sample in the sample set of marked vehicle risk parameter information contains actual vehicle risk parameter information and influence factor information, and the influence factor information comprises current value information, service life information, historical risk record information, region range information where the vehicle frequently runs and driving behavior information of a driver of a target vehicle;
Dividing the sample set of the marked risk parameter information into a training sample set and a verification sample set according to a preset proportion.
3. The integrated learning model-based risk parameter prediction method according to claim 1, wherein before the step of inputting the prediction sample set into a trained integrated learning model to perform risk parameter prediction, the method further comprises:
acquiring an initial weight vector set for each trained basic learning model;
according to the initial weight vector set for each trained basic learning model, weighting and combining each trained basic learning model by adopting a model stacking mode so as to generate a super learning model;
inputting the verification sample set into the super learning model for iterative verification, and obtaining an output result of the super learning model during each iterative verification;
acquiring the loss degree of the corresponding output result in each iteration verification relative to the risk parameter information of the vehicle risk contained in each sample in the verification sample set through a preset loss function;
until iteration verification is finished, selecting a weight vector corresponding to each basic learning model when the loss degree is the minimum value as the weight vector of the corresponding basic learning model in the super learning model, and finishing tuning of the super learning model, wherein the condition for finishing the iteration verification comprises reaching a preset maximum iteration number or the loss degree is smaller than a preset expected loss value;
And setting the super learning model with the optimized tuning as the trained integrated learning model.
4. A vehicle risk parameter prediction method based on an ensemble learning model as set forth in claim 3, characterized in that before said step of obtaining an initial weight vector set for each trained base learning model is performed, the method further includes:
different machine learning algorithms are adopted to pre-construct a plurality of basic learning models;
the k-fold cross verification mode is adopted, the training sample set is sequentially input into the plurality of basic learning models, basic learning model training is carried out, and each trained basic learning model is obtained, and the specific implementation mode is as follows:
step 401, dividing the training sample set into N training sample subsets in an equal ratio, wherein N is a positive integer;
step 402, randomly screening any training sample subset to be used as a basic verification set;
step 403, inputting other non-screened training sample subsets into a current basic learning model, and training the current basic learning model according to a built-in machine learning algorithm in the current basic learning model to obtain a pre-trained basic learning model;
Step 404, inputting the basic verification set into the pre-trained basic learning model, and performing verification tuning processing on the pre-trained basic learning model to obtain a tuned basic learning model;
step 405, using the tuned basic learning model as a currently trained basic learning model;
step 406, sequentially screening a basic learning model from the pre-constructed plurality of basic learning models as the current basic learning model, and circularly executing steps 402 to 405 to obtain each trained basic learning model.
5. The method for predicting risk parameters of vehicle risk based on an ensemble learning model as set forth in claim 4, wherein said inputting the basic verification set into the pre-trained basic learning model, performing verification tuning processing on the pre-trained basic learning model, and obtaining a tuned basic learning model, specifically includes:
step 501, inputting the basic verification set into the pre-trained basic learning model, and obtaining risk parameter information of the vehicle risk corresponding to the basic verification set output by the pre-trained basic learning model according to the influence factor information in the basic verification set;
Step 502, comparing the risk parameter information of the vehicle risk corresponding to the basic verification set output by the pre-trained basic learning model with the risk parameter information of the marked vehicle risk corresponding to the basic verification set, and obtaining a similarity comparison result;
step 503, judging whether the similarity comparison result meets a preset similarity threshold;
step 504, if the similarity comparison result does not meet the preset similarity threshold, adjusting the weight value of each influence factor in the pre-trained basic learning model, and re-executing steps 501 to 503 after adjustment;
and step 505, completing verification tuning processing of the pre-trained basic learning model until the similarity comparison result meets a preset similarity threshold value, and obtaining a tuned basic learning model.
6. The vehicle risk parameter prediction method based on an ensemble learning model as set forth in claim 3, wherein the step of inputting the verification sample set into the super learning model for iterative verification to obtain an output result of the super learning model during each iterative verification specifically includes:
inputting the verification sample set into the super learning model, and obtaining risk parameter information of the vehicle risk output by the super learning model according to the influence factor information corresponding to each sample in the verification sample set;
And taking the risk parameter information of the vehicle risk output by the super learning model as an output result of the super learning model when iterative verification is carried out at the current time.
7. The method for predicting risk parameters of vehicle risk based on an ensemble learning model according to claim 1, wherein the step of inputting the prediction sample set into a trained ensemble learning model to predict risk parameters of vehicle risk and obtain a prediction result specifically includes:
inputting the prediction sample set into a trained integrated learning model;
the integrated learning model carries out model prediction according to the influence factor information corresponding to each sample in the prediction sample set to obtain the risk parameter information of the vehicle risk corresponding to each sample output by the integrated learning model;
and taking the risk parameter information of the car risk corresponding to each sample output by the integrated learning model as the prediction result.
8. The utility model provides a vehicle risk parameter prediction device based on integrated learning model which characterized in that includes:
the prediction sample set acquisition module is used for acquiring a prediction sample set, wherein the prediction sample set is a sample set without marked risk parameter information of the vehicle risk, and each sample in the prediction sample set contains influence factor information for predicting the risk parameter of the vehicle risk, wherein the influence factor information comprises current value information, service life information, history risk record information, region range information for the vehicle to travel frequently and driving behavior information of a driver of a target vehicle;
The integrated learning model prediction module is used for inputting the prediction sample set into a trained integrated learning model to predict risk parameters of the vehicle to obtain a prediction result;
the prediction result acquisition module is used for acquiring the risk parameters of the vehicle risk of each target vehicle in the prediction sample set according to the vehicle distinguishing identification information and the prediction result, wherein the vehicle distinguishing identification information comprises license plate numbers or frame number information.
9. A computer device comprising a memory and a processor, the memory having stored therein computer readable instructions which when executed implement the steps of the ensemble learning model-based risk parameter prediction method of any one of claims 1 to 7.
10. A computer readable storage medium having stored thereon computer readable instructions which when executed by a processor implement the steps of the ensemble learning model based risk parameter prediction method of any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310741089.6A CN116777642A (en) | 2023-06-20 | 2023-06-20 | Vehicle risk parameter prediction method and device based on ensemble learning model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310741089.6A CN116777642A (en) | 2023-06-20 | 2023-06-20 | Vehicle risk parameter prediction method and device based on ensemble learning model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116777642A true CN116777642A (en) | 2023-09-19 |
Family
ID=88012879
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310741089.6A Pending CN116777642A (en) | 2023-06-20 | 2023-06-20 | Vehicle risk parameter prediction method and device based on ensemble learning model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116777642A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117150276A (en) * | 2023-11-01 | 2023-12-01 | 宁德时代新能源科技股份有限公司 | Machine learning model construction method, vehicle driving risk prediction method and device |
-
2023
- 2023-06-20 CN CN202310741089.6A patent/CN116777642A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117150276A (en) * | 2023-11-01 | 2023-12-01 | 宁德时代新能源科技股份有限公司 | Machine learning model construction method, vehicle driving risk prediction method and device |
| CN117150276B (en) * | 2023-11-01 | 2024-04-09 | 宁德时代新能源科技股份有限公司 | Machine learning model construction method, vehicle driving risk prediction method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112418059B (en) | Emotion recognition method and device, computer equipment and storage medium | |
| CN116822651A (en) | Large model parameter fine adjustment method, device, equipment and medium based on incremental learning | |
| CN116777642A (en) | Vehicle risk parameter prediction method and device based on ensemble learning model | |
| CN111967581A (en) | Interpretation method and device of clustering model, computer equipment and storage medium | |
| CN116934283A (en) | Employee authority configuration method, device, equipment and storage medium thereof | |
| CN116777646A (en) | Artificial intelligence-based risk identification method, apparatus, device and storage medium | |
| CN116383787A (en) | Page creation method, page creation device, computer equipment and storage medium | |
| CN113902576A (en) | Deep learning-based information pushing method and device, electronic equipment and medium | |
| CN112257812A (en) | Method and device for determining labeled sample, machine readable medium and equipment | |
| CN116307742B (en) | Risk identification method, device and equipment for subdivision guest group and storage medium | |
| CN114238583B (en) | Natural language processing method, device, computer equipment and storage medium | |
| CN115640896B (en) | Household user power load prediction method under multi-user scene and related equipment | |
| CN116777645A (en) | Method for enhancing and optimizing risk prediction model of vehicle risk and related equipment thereof | |
| CN116797381A (en) | Health risk underwriting risk prediction method and related equipment thereof | |
| CN117252712A (en) | Product claim settlement method, device, equipment and storage medium based on transfer learning | |
| CN117829899A (en) | Demand prediction method, device, equipment and storage medium based on artificial intelligence | |
| CN116681534A (en) | Vehicle risk fraud prediction method based on improved DTW algorithm and related equipment thereof | |
| CN117034114A (en) | Data prediction method, device, equipment and storage medium based on artificial intelligence | |
| CN116993516A (en) | Interpretation optimization method, device and equipment for claim settlement model and storage medium thereof | |
| CN116702995A (en) | Vehicle risk index prediction method and related equipment thereof | |
| CN116823499A (en) | Model-based vehicle insurance pricing method, system, computer equipment and storage medium | |
| CN117235633A (en) | Mechanism classification method, mechanism classification device, computer equipment and storage medium | |
| CN117395253A (en) | Load balancing method, device, equipment and storage medium thereof | |
| CN113902575A (en) | Deep learning-based information pushing method and device, electronic equipment and medium | |
| CN117829961A (en) | Financial product recommendation method, device and computer equipment based on questionnaire model |
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 |