Vehicle owner-oriented credit score prediction method, system and device, and storage medium

Abstract

The invention relates to a vehicle owner-oriented credit score prediction method, system and device and a storage medium, and the method comprises the steps: obtaining multi-dimensional user score data, and employing an automatic feature extraction algorithm to extract multi-dimensional key features from the multi-dimensional user score data according to the correlation between user behaviors and credit risks; inputting the multi-dimensional key features into a credit score prediction model to obtain a user credit score; and performing user credit risk layering according to the user credit score, capturing a change condition of the user credit score, dynamically adjusting a layering strategy according to the change condition, and generating a personalized financial service strategy for users of different levels. The credit scoring method has the effect of improving the credit scoring accuracy.

Description

Technical Field

[0001] The present application relates to the technical field of credit scoring, and in particular to a credit score prediction method, system, device and storage medium for car owners. Background Art

[0002] At present, with the development of the financial market and the increase in demand for personalized services, credit scoring is increasingly used in financial services. However, most existing credit scoring methods rely on users' historical credit records and limited financial behavior data, lacking a comprehensive analysis of users' daily behavior, especially the credit risk assessment of car owners. Since data such as car owners' vehicle usage behavior, payment behavior, and geographic location are not effectively incorporated into existing credit scoring models, these potential risk factors are not fully utilized, resulting in insufficient accuracy and real-time performance of credit scoring.

[0003] Existing credit scoring models mostly use static data input and are unable to dynamically adjust the scoring results based on the user's real-time behavior, causing the credit score to lag behind the user's actual risk status. This is especially true for car owners whose behavior changes frequently. Existing models often cannot capture their credit risk dynamics in a timely manner. In addition, key factors such as abnormal vehicle usage patterns and abnormal fluctuations in payment frequency have not been fully considered in existing credit scoring models, resulting in limited risk prediction capabilities.

[0004] The above-mentioned existing technical solutions have the following defects: the existing credit scoring method mainly relies on static historical data, resulting in low accuracy of credit scoring, so there is room for improvement. Summary of the invention

[0005] In order to improve the accuracy of credit scores, the present application provides a credit score prediction method, system, device and storage medium for car owners.

[0006] The above-mentioned invention objective of the present application is achieved through the following technical solutions: A credit score prediction method for a car owner, the credit score prediction method for a car owner comprising: Obtain multi-dimensional user rating data, and extract multi-dimensional key features from the multi-dimensional user rating data using an automatic feature extraction algorithm based on the correlation between user behavior and credit risk; Inputting the multi-dimensional key features into a credit score prediction model to obtain a user credit score; Perform user credit risk stratification based on the user credit score, capture changes in the user credit score, dynamically adjust the stratification strategy based on the changes, and generate personalized financial service strategies for users of different levels.

[0007] By adopting the above technical solution, by obtaining multi-dimensional user rating data, a comprehensive information source can be provided for credit scoring, thereby improving the accuracy and comprehensiveness of the scoring results; by extracting multi-dimensional key features from multi-dimensional user rating data based on the correlation between user behavior and credit risk, an automatic feature extraction algorithm can be used to effectively filter redundant data and extract features that are highly correlated with credit risk, thereby improving the efficiency and accuracy of the model; by inputting multi-dimensional key features into a credit scoring prediction model to obtain a user credit score, a personalized credit score can be calculated based on the user's current behavior data, thereby improving the timeliness and accuracy of the credit score; by performing user credit risk stratification based on user credit scores and capturing changes in user credit scores, dynamically adjusting stratification strategies based on changes, and generating personalized financial service strategies for users at different levels, personalized services can be provided based on the credit risk conditions of different users, thereby enhancing the risk management capabilities and user experience of financial institutions.

[0008] In one example, the present application may be further configured as follows: the multi-dimensional user rating data is obtained, and according to the correlation between user behavior and credit risk, multi-dimensional key features are extracted from the multi-dimensional user rating data using an automatic feature extraction algorithm, specifically including: Acquire vehicle usage data, payment behavior data, geographic location information, historical credit data, etc. as the user rating data, and analyze the time series correlation of the user rating data; Using a time series analysis method, analyzing the time series correlation of the user rating data to obtain the correlation between the user rating and the credit risk, and sorting the user rating data according to the credit risk to obtain a user rating arrangement set; The user rating arrangement set is reduced in dimension based on the automatic feature extraction algorithm to obtain the multi-dimensional key features.

[0009] By adopting the above technical solution, by obtaining vehicle usage data, payment behavior data, geographic location information and historical credit data as user scoring data, and analyzing the time series correlation of user scoring data, it is possible to provide more accurate time series data input for the credit scoring model, thereby improving the reliability of risk assessment; by adopting a time series analysis method, according to the time series correlation of user scoring data, analysis is performed to obtain the correlation between user scoring and credit risk, and user scoring data is sorted according to credit risk to obtain a user scoring permutation set, which can better reflect the relationship between user behavior patterns and risks, thereby improving the risk perception ability of the model; by reducing the dimension of the user scoring permutation set based on an automatic feature extraction algorithm, multi-dimensional key features are obtained, which can simplify the data dimension and retain key features, thereby improving the operational efficiency and accuracy of the model.

[0010] In one example, the present application may be further configured as follows: the construction of the credit score prediction model specifically includes: Obtaining user historical rating data, and performing correlation analysis on the historical user rating data using a time series analysis algorithm to obtain corresponding historical rating credit risks, and classifying the user historical rating data according to the historical rating credit risks to obtain a model training data set; The preset machine learning algorithm is iteratively trained using the model training data set, and the impact of different data in the user's historical scoring data on the score change is analyzed according to the historical score credit risk to obtain the credit score prediction model.

[0011] By adopting the above technical solution, by obtaining the user's historical scoring data and using the time series analysis algorithm to perform correlation analysis, it is possible to identify key change patterns related to credit risk based on the user's historical scoring records, thereby helping the system to establish different risk categories and improve the prediction basis of the model; by using the layered and classified model training data set to iteratively train the preset machine learning algorithm, the model can accurately learn the impact of different behavioral characteristics on credit scores, thereby improving the model's prediction accuracy and flexibility for future credit scores. This overall process realizes deep learning based on user historical data, so that the credit scoring model can not only adapt to user behavior changes in real time, but also identify potential risks more accurately, providing financial institutions with a more comprehensive and reliable user credit risk assessment.

[0012] In one example, the present application may be further configured as follows: the construction of the credit score prediction model further includes: Dynamically adjust the feature weights in the model based on changes in user behavior patterns to improve the model's adaptability and assessment accuracy to credit risk.

[0013] By adopting the above technical solution, the feature weights in the model are dynamically adjusted according to changes in user behavior patterns to improve the model's adaptability to credit risk and assessment accuracy. This can respond to changes in user behavior in real time, increase the model's adaptability to credit risk, and thus improve the flexibility and accuracy of risk prediction.

[0014] In one example, the present application may be further configured as follows: performing user credit risk stratification according to the user credit score specifically includes: According to the association between the vehicle usage data, the payment behavior data and the user credit score, a change signal of the user's credit status is captured, and the level of the credit risk stratification is determined according to the change signal; The user with abnormal vehicle usage data is marked as high risk, and the user credit risk stratification is performed according to the credit risk stratification level and / or the high risk mark.

[0015] By adopting the above technical solution, by capturing the change signals of the user's credit status based on the correlation between the vehicle usage data, payment behavior data and the user's credit score, and determining the level of credit risk stratification based on the change signals, it is possible to timely identify changes in the user's credit status, thereby dynamically adjusting the risk stratification and enhancing the ability to manage user risks; by marking users with abnormal vehicle usage data as high-risk, and performing user credit risk stratification based on the credit risk stratification level and / or high-risk marking, it is possible to effectively identify high-risk users, and provide more accurate risk control based on risk stratification, thereby reducing the risk exposure of financial institutions.

[0016] In one example, the present application may be further configured as follows: capturing the change of the user's credit score and dynamically adjusting the tiering strategy according to the change specifically includes: Automatically update the user's credit score based on the real-time user score data; In the event of a significant fluctuation in the credit score, the tiering strategy is adjusted in real time by analyzing the abnormalities in the vehicle usage frequency and the payment behavior data.

[0017] By adopting the above technical solution, the user's credit score is automatically updated based on real-time user rating data, which can reflect the user's latest behavioral changes in real time, ensure the timeliness and accuracy of the credit score, and thus provide a reliable basis for financial decision-making; by analyzing the abnormalities of vehicle usage frequency and payment behavior data in the event of large fluctuations in credit scores, the stratification strategy can be adjusted in real time, which can flexibly respond to changes in user credit risk and avoid outdated or inaccurate risk assessments, thereby improving the dynamic adjustment capability and accuracy of risk stratification.

[0018] The second object of the invention is achieved by the following technical solutions: A credit score prediction system for a car owner, the credit score prediction system for a car owner comprising: Obtain multi-dimensional user rating data, and extract multi-dimensional key features from the multi-dimensional user rating data using an automatic feature extraction algorithm based on the correlation between user behavior and credit risk; Inputting the multi-dimensional key features into a credit score prediction model to obtain a user credit score; Perform user credit risk stratification based on the user credit score, capture changes in the user credit score, dynamically adjust the stratification strategy based on the changes, and generate personalized financial service strategies for users of different levels.

[0019] By adopting the above technical solution, by obtaining multi-dimensional user rating data, a comprehensive information source can be provided for credit scoring, thereby improving the accuracy and comprehensiveness of the scoring results; by extracting multi-dimensional key features from multi-dimensional user rating data based on the correlation between user behavior and credit risk, an automatic feature extraction algorithm can be used to effectively filter redundant data and extract features that are highly correlated with credit risk, thereby improving the efficiency and accuracy of the model; by inputting multi-dimensional key features into a credit scoring prediction model to obtain a user credit score, a personalized credit score can be calculated based on the user's current behavior data, thereby improving the timeliness and accuracy of the credit score; by performing user credit risk stratification based on user credit scores and capturing changes in user credit scores, dynamically adjusting stratification strategies based on changes, and generating personalized financial service strategies for users at different levels, personalized services can be provided based on the credit risk conditions of different users, thereby enhancing the risk management capabilities and user experience of financial institutions.

[0020] The third objective of the present application is achieved through the following technical solutions: A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned credit score prediction method for vehicle owners when executing the computer program.

[0021] The fourth objective of the present application is achieved through the following technical solutions: A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned credit score prediction method for car owners.

[0022] In summary, this application includes the following beneficial technical effects: 1. By acquiring multi-dimensional user rating data, it can provide a comprehensive source of information for credit rating, thereby improving the accuracy and comprehensiveness of the rating results; by extracting multi-dimensional key features from multi-dimensional user rating data based on the correlation between user behavior and credit risk, using an automatic feature extraction algorithm, it can effectively filter redundant data and extract features that are highly correlated with credit risk, thereby improving the efficiency and accuracy of the model; by inputting multi-dimensional key features into the credit rating prediction model to obtain user credit scores, it can calculate personalized credit scores based on the user's current behavior data, thereby improving the timeliness and accuracy of credit ratings; by performing user credit risk stratification based on user credit scores and capturing changes in user credit scores, dynamically adjusting stratification strategies based on changes, and generating personalized financial service strategies for users at different levels, it can provide personalized services based on the credit risk conditions of different users, thereby improving the risk management capabilities and user experience of financial institutions; 2. By obtaining vehicle usage data, payment behavior data, geographic location information and historical credit data as user scoring data and analyzing the time series correlation of user scoring data, it is possible to provide more accurate time series data input for the credit scoring model, thereby improving the reliability of risk assessment; by adopting the time series analysis method, according to the time series correlation of user scoring data, the correlation between user scoring and credit risk is obtained, and the user scoring data is sorted according to credit risk to obtain the user scoring permutation set, which can better reflect the relationship between user behavior patterns and risks, thereby improving the risk perception ability of the model; by reducing the dimension of the user scoring permutation set based on the automatic feature extraction algorithm, multi-dimensional key features are obtained, which can simplify the data dimension and retain the key features, thereby improving the operational efficiency and accuracy of the model; 3. By obtaining the user's historical behavior characteristics and using machine learning algorithms to iteratively train the user's historical behavior characteristics, the model's ability to learn the user's historical behavior patterns can be enhanced, thereby improving the prediction accuracy of the credit score; by introducing a vehicle usage abnormal pattern detection module, the score correction is performed for vehicle usage behaviors with abnormal frequencies, and a credit score prediction model is obtained, which can effectively identify the user's abnormal behavior and dynamically adjust the credit score, thereby improving the model's sensitivity to risk events and real-time response capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a flow chart of a credit score prediction method for a car owner in one embodiment of the present application; Figure 2 is a flowchart for implementing step S10 in a credit score prediction method for car owners in one embodiment of the present application; Figure 3It is a flowchart for implementing the construction of a credit score prediction model in a credit score prediction method for car owners in one embodiment of the present application; Figure 4 It is another implementation flow chart of constructing a credit score prediction model in a credit score prediction method for car owners in one embodiment of the present application; Figure 5 , Figure 6 is a flowchart for implementing step S30 in a credit score prediction method for car owners in one embodiment of the present application; Figure 7 It is a principle block diagram of a credit score prediction system for car owners in one embodiment of the present application; Figure 8 It is a schematic diagram of a device in an embodiment of the present application. DETAILED DESCRIPTION

[0024] The present application is further described in detail below in conjunction with the accompanying drawings.

[0025] In one embodiment, if Figure 1 As shown, the present application discloses a credit score prediction method for car owners, which specifically includes the following steps: S10: Obtain multi-dimensional user rating data, and use an automatic feature extraction algorithm to extract multi-dimensional key features from the multi-dimensional user rating data based on the correlation between user behavior and credit risk.

[0026] Specifically, the analysis is performed by integrating user rating data from multiple dimensions, such as vehicle usage data, payment behavior data, geographic location information, and historical credit data. All collected data is filtered and formatted in a preprocessing module, and further extracted through an automatic feature extraction algorithm. Vehicle usage data, including data such as mileage, usage frequency, and maintenance records, infers the user's behavioral risk by analyzing the vehicle's usage patterns and frequency. Payment behavior data, including payment amount, payment method, payment frequency, etc., identifies potential risks through analysis of payment habits and large payments. At the same time, geographic location information is combined with the user's range of activities and frequented places to further enhance the understanding of behavioral patterns, and historical credit data provides the user's historical credit performance. All of these feature data are screened according to relevance through a feature extraction algorithm and input into the subsequent credit score prediction model in a weighted form.

[0027] S20: Input the multi-dimensional key features into the credit score prediction model to obtain the user credit score.

[0028] Specifically, the extracted multi-dimensional key features will be input into the pre-trained credit score prediction model, which combines machine learning algorithms such as logistic regression and XGBoost to predict the user's credit risk. The model assigns weights to each input feature and adjusts the score based on the performance of different features in historical data. When performing predictions, the model first assigns and calculates the weights to each feature through the trained weights, and finally generates the user's credit score, which is a comprehensive evaluation based on the current user behavior pattern and historical behavior performance, providing a basis for the subsequent generation of personalized financial service strategies.

[0029] S30: Perform user credit risk stratification based on user credit scores, capture changes in user credit scores, dynamically adjust stratification strategies based on changes, and generate personalized financial service strategies for users of different levels.

[0030] Specifically, based on the currently calculated credit score, users will be assigned to different credit risk levels. Credit risk stratification is based on the dynamic performance of users' credit scores and historical behavioral data, and further refines credit risks based on vehicle usage patterns, payment behaviors, and geographic location information. The system continuously monitors changes in user behavior, such as vehicle usage frequency or abnormal payment fluctuations, captures fluctuations in user credit scores in real time, and updates their risk levels through tiered adjustment strategies when user risks change. For users with high credit ratings, more favorable loan interest rates, higher credit limits, value-added services and other personalized financial service strategies are provided; for users with large fluctuations in credit scores or who are at high risk levels, measures such as reducing credit limits and increasing risk control may be taken. All strategies are based on users' dynamic behavioral performance to ensure the accuracy and flexibility of services.

[0031] In one embodiment, if Figure 2 As shown, in step S10, multi-dimensional user rating data is obtained, and according to the correlation between user behavior and credit risk, multi-dimensional key features are extracted from the multi-dimensional user rating data using an automatic feature extraction algorithm, specifically including: S11: Obtain vehicle usage data, payment behavior data, geographic location information, historical credit data, etc. as user rating data, and analyze the time series correlation of the user rating data.

[0032] Specifically, multi-dimensional data is collected in real time through Internet of Vehicles data, payment systems, and credit record systems, including data such as vehicle mileage, usage frequency, payment behavior, and the user's geographic activity area. The system will perform time series analysis on various types of data to determine the degree of correlation between vehicle usage behavior and payment behavior and user credit risk, and identify regular changes in user behavior. Through time series correlation analysis, it can be found out whether the user's high-frequency payment behavior is related to the increase in vehicle usage frequency, thereby accurately predicting potential credit risks.

[0033] S12: Using a time series analysis method, the temporal correlation of the user rating data is analyzed to obtain the correlation between the user rating and the credit risk, and the user rating data is sorted according to the credit risk to obtain a user rating arrangement set.

[0034] Specifically, the collected user rating data is processed through time series analysis, and combined with correlation analysis tools to evaluate the impact of various behavior patterns on user credit risk. The analysis process not only considers the user's payment behavior and vehicle usage data within a specific period of time, but also combines historical credit behavior records to generate a comprehensive risk assessment report. For each user behavior pattern, the system will sort it according to its correlation with credit risk, and generate a user rating permutation set based on these rankings. The user rating permutation set will be further used for modeling and credit score calculation to ensure the accuracy and effectiveness of the final credit risk assessment.

[0035] S13: Based on the automatic feature extraction algorithm, the user rating arrangement set is reduced in dimension to obtain multi-dimensional key features.

[0036] Specifically, the user rating permutation set is processed by dimensionality reduction through an automatic feature extraction algorithm. Dimensionality reduction algorithms such as principal component analysis will simplify high-dimensional data while retaining key features that are highly relevant to credit risk assessment. During the dimensionality reduction process, the system will reduce the dimension of the data by removing redundant data, reducing data complexity and other operations to improve computing efficiency. At the same time, the retained feature sets such as vehicle usage frequency and abnormal payment fluctuations will become the main basis for subsequent credit scoring model input. Through the key features after dimensionality reduction processing, the system can perform risk assessment more efficiently, reduce the computational burden of the model, and improve the accuracy of predictions.

[0037] In one embodiment, if Figure 3 As shown, in step S20, i.e., the construction of the credit score prediction model, specifically includes: S201: Obtain user historical rating data, and use a time series analysis algorithm to perform correlation analysis on the historical user rating data to obtain the corresponding historical rating credit risk, classify the user historical rating data according to the historical rating credit risk, and obtain a model training data set.

[0038] Specifically, the user's historical scoring data is first extracted from the historical database. The data contains the user's credit score records and related behavioral data in different time periods. Through the time series analysis algorithm, these historical scoring data are deeply analyzed to identify the correlation between score changes between different time points. Time series analysis can help identify different patterns such as long-term trends, seasonal fluctuations and short-term fluctuations. Combined with abnormal situations in score changes, such as time points when scores drop or rise sharply, the system can generate corresponding historical score credit risks. These risk data are used to stratify and classify the user's historical scoring data, and divide users into multiple categories according to the risk level of their historical score changes, thereby forming model training data sets of different categories. The training data set will be used for subsequent machine learning model training so that the model can learn the relationship between score changes and risks.

[0039] S202: Iteratively train the preset machine learning algorithm using the model training data set, analyze the impact of different data in the user's historical scoring data on the score change based on the historical score credit risk, and obtain a credit score prediction model.

[0040] Specifically, in the model training phase, the model training data set is input into a preset machine learning algorithm such as XGBoost, random forest or logistic regression, and multiple iterations of training are performed. In each iteration, the model will perform stratified learning on the historical score credit risks in the data set through the cross-validation method, and automatically adjust the weights of the features to identify the specific impact of different data features on score changes. For example, the model will analyze the relationship between fluctuations in vehicle usage frequency and score changes, or the relationship between abnormal payment behavior and score fluctuations. During the iterative training process, the model assigns corresponding weights according to the importance of different features to improve the model's prediction accuracy and flexibility. The credit score prediction model that finally completes the training can identify the impact of different behavioral characteristics on user credit scores and be used in future score prediction tasks.

[0041] In one embodiment, if Figure 4 As shown, in step S20, i.e., the construction of the credit score prediction model, it also includes: S2021: Dynamically adjust the feature weights in the model based on changes in user behavior patterns to improve the model's adaptability and assessment accuracy to credit risk.

[0042] Specifically, users' behavior patterns will change dynamically during their daily vehicle use and payment. To ensure that the credit scoring model can accurately reflect the user's risk status, the model will continuously monitor these behavioral characteristics and adjust the feature weights in real time according to their changes. For example, when the user's payment behavior increases abnormally or the frequency of vehicle use suddenly increases, the model will increase the weights of these features, thereby increasing their influence in the credit score. Through this dynamic adjustment mechanism, the system ensures that the model is not only suitable for the analysis of static data, but can also adapt to real-time behavioral changes of users, improve the model's accurate assessment of risks, and help financial institutions better identify and manage potential risks.

[0043] In one embodiment, if Figure 5 As shown, in step S30, user credit risk stratification is performed according to the user credit score, specifically including: S31: Based on the correlation between vehicle usage data, payment behavior data and user credit score, capture the change signal of the user's credit status, and determine the level of credit risk stratification based on the change signal.

[0044] Specifically, there is a direct correlation between vehicle usage data, payment behavior data and the user's credit score. Through continuous monitoring of these data, the system can capture changes in the user's credit status. For example, when the frequency of vehicle use increases and occurs simultaneously with large-amount payment behaviors, this may be a signal of rising credit risk. When the system detects these change signals, it will analyze the user's current credit score, determine in real time whether his or her credit risk level needs to be adjusted, and execute user risk stratification. This risk stratification mechanism based on behavioral changes enables the system to respond to changes in user credit risk more quickly and accurately.

[0045] S32: Mark users with abnormal vehicle usage data as high risk, and perform user credit risk stratification according to the credit risk stratification level and / or high risk mark.

[0046] Specifically, when abnormalities appear in vehicle usage data, such as a user's vehicle usage frequency rises sharply in a short period of time or the mileage increases significantly, the system will automatically mark the user as high-risk. This high-risk mark can help the system identify potential risky users, especially in combination with abnormal fluctuations in payment behavior and credit scores. Once the high-risk mark is triggered, the system will re-execute the user's risk stratification based on the user's current credit risk stratification level and the prompt of the high-risk mark, and may increase the user's credit risk level. This stratification adjustment mechanism helps to respond to potential credit risks in a timely manner and ensure that the system's risk management of high-risk users is more accurate and effective.

[0047] In one embodiment, if Figure 6As shown, in step S30, the changes in the user's credit score are captured, and the tiering strategy is dynamically adjusted according to the changes, specifically including: S33: Automatically update the user's credit score based on real-time user rating data.

[0048] Specifically, the user's credit score will be automatically updated as the behavioral data changes dynamically. The system obtains key data such as the user's payment behavior, vehicle usage frequency, and geographic location changes in real time, and analyzes this data through a credit scoring model. If the user's behavior is stable and the credit risk is low, the system will keep their credit score unchanged or slightly increase it; conversely, if the user is detected to have abnormal payment behavior or vehicle usage, the system will automatically adjust their credit score. This dynamic update mechanism based on real-time data ensures that the credit score can accurately reflect the user's current credit status and avoids the score lagging behind the actual risk.

[0049] S34: In the event of significant fluctuations in credit scores, the tiering strategy is adjusted in real time by analyzing anomalies in vehicle usage frequency and payment behavior data.

[0050] Specifically, when the system detects a significant fluctuation in the user's credit score, it will first conduct a detailed analysis of the user's vehicle usage data and payment behavior. The system will check whether there is a significant change in the frequency of vehicle use and whether the payment behavior deviates significantly from the user's historical records. If it is confirmed that these abnormal behaviors are correlated with fluctuations in credit scores, the system will adjust the user's risk stratification strategy in real time based on actual conditions and re-evaluate the user's credit risk level. This real-time adjustment mechanism ensures that the user's risk stratification strategy can flexibly respond to emergencies and avoid decision-making delays caused by score fluctuations.

[0051] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0052] In one embodiment, a credit score prediction system for car owners is provided, and the credit score prediction system for car owners corresponds one-to-one with the credit score prediction method for car owners in the above embodiment. Figure 7 As shown, the credit score prediction system for car owners includes a data collection module, a credit score calculation module, and a credit risk stratification module. The detailed description of each functional module is as follows: The data collection module is used to obtain multi-dimensional user rating data and extract multi-dimensional key features from the multi-dimensional user rating data using an automatic feature extraction algorithm based on the correlation between user behavior and credit risk; The credit score calculation module is used to input multi-dimensional key features into the credit score prediction model to obtain the user's credit score; The credit risk stratification module is used to perform user credit risk stratification based on user credit scores, capture changes in user credit scores, dynamically adjust stratification strategies based on changes, and generate personalized financial service strategies for users of different levels.

[0053] Optionally, the data acquisition module specifically includes: The data processing submodule is used to obtain vehicle usage data, payment behavior data, geographic location information, and historical credit data as user rating data, and analyze the temporal correlation of user rating data; The time series analysis submodule is used to use the time series analysis method to analyze the time series correlation of user rating data to obtain the correlation between user rating and credit risk, and to sort the user rating data according to credit risk to obtain a user rating arrangement set; The feature dimension reduction submodule is used to reduce the dimension of the user rating arrangement set based on the automatic feature extraction algorithm to obtain multi-dimensional key features.

[0054] Optionally, the construction of the credit score prediction model specifically includes: The data acquisition module is used to obtain the user's historical rating data, and use the time series analysis algorithm to perform correlation analysis on the historical user rating data to obtain the corresponding historical rating credit risk, and classify the user's historical rating data according to the historical rating credit risk to obtain the model training data set; The model training module is used to iteratively train the preset machine learning algorithm using the model training data set, analyze the impact of different data in the user's historical scoring data on the score changes based on the historical scoring credit risk, and obtain a credit score prediction model.

[0055] Optionally, the construction of the credit score prediction model also includes: The feature weight adjustment submodule is used to dynamically adjust the feature weights in the model according to changes in user behavior patterns to improve the model's adaptability to credit risk and assessment accuracy.

[0056] Optionally, the credit risk stratification module specifically includes: The credit status capture submodule is used to capture the change signal of the user's credit status based on the association between the vehicle usage data, payment behavior data and the user's credit score, and determine the level of credit risk stratification based on the change signal; The high-risk marking submodule is used to mark users with abnormal vehicle usage data as high-risk and perform user credit risk stratification based on the credit risk stratification level and / or high-risk marking.

[0057] The credit score update submodule is used to automatically update the user's credit score based on real-time user score data; The tiered strategy adjustment submodule is used to adjust the tiered strategy in real time by analyzing the abnormalities in vehicle usage frequency and payment behavior data when there are significant fluctuations in credit scores.

[0058] For the specific limitations of the credit score prediction system for car owners, please refer to the limitations of the credit score prediction method for car owners mentioned above, which will not be repeated here. Each module in the above-mentioned credit score prediction system for car owners can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0059] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a credit score prediction method for a car owner is implemented.

[0060] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: Obtain multi-dimensional user rating data, and use automatic feature extraction algorithms to extract multi-dimensional key features from the multi-dimensional user rating data based on the correlation between user behavior and credit risk; Input the multi-dimensional key features into the credit score prediction model to obtain the user's credit score; Perform user credit risk stratification based on user credit scores, capture changes in user credit scores, dynamically adjust stratification strategies based on changes, and generate personalized financial service strategies for users of different levels.

[0061] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: Obtain multi-dimensional user rating data, and use automatic feature extraction algorithms to extract multi-dimensional key features from the multi-dimensional user rating data based on the correlation between user behavior and credit risk; Input the multi-dimensional key features into the credit score prediction model to obtain the user's credit score; Perform user credit risk stratification based on user credit scores, capture changes in user credit scores, dynamically adjust stratification strategies based on changes, and generate personalized financial service strategies for users of different levels.

[0062] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0063] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above.

[0064] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A credit score prediction method for car owners, characterized in that: The credit score prediction method for the car owner includes: Obtain multi-dimensional user rating data, and extract multi-dimensional key features from the multi-dimensional user rating data using an automatic feature extraction algorithm based on the correlation between user behavior and credit risk; Inputting the multi-dimensional key features into a credit score prediction model to obtain a user credit score; Perform user credit risk stratification based on the user credit score, capture changes in the user credit score, dynamically adjust the stratification strategy based on the changes, and generate personalized financial service strategies for users of different levels.

2. The credit score prediction method for car owners according to claim 1, characterized in that: The multi-dimensional user rating data is obtained, and based on the correlation between user behavior and credit risk, an automatic feature extraction algorithm is used to extract multi-dimensional key features from the multi-dimensional user rating data, specifically including: Acquire vehicle usage data, payment behavior data, geographic location information, historical credit data, etc. as the user rating data, and analyze the time series correlation of the user rating data; Using a time series analysis method, analyzing the time series correlation of the user rating data to obtain the correlation between the user rating and the credit risk, and sorting the user rating data according to the credit risk to obtain a user rating arrangement set; The user rating arrangement set is reduced in dimension based on the automatic feature extraction algorithm to obtain the multi-dimensional key features.

3. The credit score prediction method for car owners according to claim 1, characterized in that: The construction of the credit score prediction model specifically includes: Obtaining user historical rating data, and performing correlation analysis on the historical user rating data using a time series analysis algorithm to obtain corresponding historical rating credit risks, and classifying the user historical rating data according to the historical rating credit risks to obtain a model training data set; The preset machine learning algorithm is iteratively trained using the model training data set, and the impact of different data in the user's historical scoring data on the score change is analyzed according to the historical score credit risk to obtain the credit score prediction model.

4. The credit score prediction method for car owners according to claim 3, characterized in that: The construction of the credit score prediction model also includes: Dynamically adjust the feature weights in the model based on changes in user behavior patterns to improve the model's adaptability and assessment accuracy to credit risk.

5. The credit score prediction method for car owners according to claim 2, characterized in that: The performing of user credit risk stratification according to the user credit score specifically includes: According to the association between the vehicle usage data, the payment behavior data and the user credit score, a change signal of the user's credit status is captured, and the level of the credit risk stratification is determined according to the change signal; The user with abnormal vehicle usage data is marked as high risk, and the user credit risk stratification is performed according to the credit risk stratification level and / or the high risk mark.

6. The credit score prediction method for car owners according to claim 5, characterized in that: The capturing of changes in the user's credit score and dynamically adjusting the tiering strategy according to the changes specifically includes: Automatically update the user's credit score based on the real-time user score data; In the event of a significant fluctuation in the credit score, the tiering strategy is adjusted in real time by analyzing the abnormalities in the vehicle usage frequency and the payment behavior data.

7. A credit score prediction system for car owners, characterized in that: The credit score prediction system for car owners includes: A data collection module is used to obtain multi-dimensional user rating data and, based on the correlation between user behavior and credit risk, extract multi-dimensional key features from the multi-dimensional user rating data using an automatic feature extraction algorithm; A credit score calculation module, used to input the multi-dimensional key features into a credit score prediction model to obtain a user credit score; The credit risk stratification module is used to perform user credit risk stratification according to the user credit score, capture changes in the user credit score, dynamically adjust the stratification strategy according to the changes, and generate personalized financial service strategies for users of different levels.

8. The credit score prediction system for car owners according to claim 7, characterized in that: The data acquisition module specifically includes: A data processing submodule, for acquiring vehicle usage data, payment behavior data, geographic location information and historical credit data as the user rating data, and analyzing the temporal correlation of the user rating data; A time series analysis submodule, for analyzing the time series correlation of the user rating data using a time series analysis method to obtain the correlation between the user rating and the credit risk, and sorting the user rating data according to the credit risk to obtain a user rating arrangement set; The feature dimension reduction submodule is used to reduce the dimension of the user rating arrangement set based on the automatic feature extraction algorithm to obtain the multi-dimensional key features.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the credit score prediction method for car owners as described in any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the credit score prediction method for car owners as claimed in any one of claims 1 to 6 are implemented.

Images