AI-based automobile financial dynamic risk assessment system and method

Abstract

The invention discloses an AI-based automobile finance dynamic risk assessment system and method, and particularly relates to the technical field of artificial intelligence. Basic information of customers, including credit score, income condition, loan history, brand, model and age limit information of vehicles, is acquired, the customer information is processed by using a machine learning algorithm, and the credit risk of the customers is assessed; the method comprises the following steps: acquiring customer loan historical data, generating a corresponding credit score, judging the repayment capability and past default records of a customer based on the customer loan historical data, evaluating the repayment risk of the customer, estimating the value of a vehicle based on the brand, model and age limit information of the vehicle in combination with the market trend, judging the mortgage risk of the vehicle, and integrating the credit score, the repayment risk and the mortgage risk. And performing deep fusion analysis on the multi-dimensional risk factors by using an AI algorithm, calculating a final comprehensive risk level, and giving suggestions on whether to approve loan, loan amount and interest rate adjustment according to a preset risk threshold.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and more specifically, to an AI-based dynamic risk assessment system and method for automotive finance. Background Technology

[0002] With the rapid development of the auto finance market, traditional auto finance approval systems face numerous challenges. Most traditional systems are based on fixed rules and preset thresholds, relying primarily on factors such as credit history, income status, and vehicle valuation for approval decisions.

[0003] However, due to the complexity of the financial market environment and the variability of customer behavior patterns, missing data, outliers, and the non-linear effects of market fluctuations can make it difficult to accurately identify risks in risk assessments. Furthermore, existing technologies have failed to effectively integrate different types of data sources. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an AI-based dynamic risk assessment system and method for automotive finance, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an AI-based dynamic risk assessment system for auto finance, comprising a customer information collection module, a customer credit assessment module, a loan history analysis module, a vehicle valuation module, a risk assessment module, and a data storage module; The customer information collection module collects basic customer information, including credit score, income status, loan history, vehicle brand, model, and age, and sends the customer information to the customer credit assessment module. The customer credit assessment module uses machine learning algorithms to process the received customer information, assess the customer's credit risk, and generate a corresponding credit score. The loan history analysis module is based on the customer's loan history data to determine the customer's repayment ability and past default records, and to assess their repayment risk. The vehicle valuation module is based on the vehicle's brand, model, and year of manufacture, combined with market trends, to value the vehicle and assess its mortgage risk. The risk assessment module integrates the credit score generated by the customer credit assessment module, the repayment risk judged by the loan history analysis module, and the mortgage risk judged by the vehicle valuation module. It uses AI algorithms to perform in-depth fusion analysis on the above multi-dimensional risk factors, calculates the final comprehensive risk level, and provides suggestions on whether to approve the loan, the loan amount, and interest rate adjustments based on preset risk thresholds. The final result is then sent to the data storage module for recording.

[0006] Preferably, as a preferred embodiment of the AI-based dynamic risk assessment system for auto finance described in this invention, the customer information collection module collects basic customer information, including credit score, income status, loan history, vehicle brand, model, and age, and sends the customer information to the customer credit assessment module, specifically including: The system collects basic identity information, personal financial information, and vehicle information of customers. The basic identity information includes name, ID number, contact information, and address. The personal financial information includes credit report data, income data, and loan history data. The vehicle information includes the brand, model, age, mileage, and VIN code information of the vehicle that the customer intends to purchase or mortgage. The collected data undergoes preliminary formatting and basic data integrity and compliance verification. The pre-processed and verified customer basic information, credit report data, income data, loan history data, and vehicle information are then securely packaged and sent to the customer credit assessment module.

[0007] Preferably, as a preferred embodiment of the AI-based dynamic risk assessment system for auto finance described in this invention, the customer credit assessment module utilizes machine learning algorithms to process received customer information, assess the customer's credit risk, and generate a corresponding credit score, specifically including: It receives basic customer identity information, personal financial information, and vehicle information from the customer information collection module, which have been preliminarily formatted and verified. It extracts key features from the received raw data, including credit report features, income and debt features, loan history features, and vehicle features, in order to maximize information utilization efficiency and adapt to machine learning models. The credit report features include historical credit scores, number of overdue payments and days, number of credit accounts, credit card limit utilization rate, number of loan applications, multiple borrowing, and number of credit inquiries; the income and debt features include average monthly income, income source type, and income-to-debt ratio; the loan history features include total historical loan amount, amount repaid, outstanding balance, loan term, and number of prepayments; the vehicle features include calculating the residual value rate of the vehicle brand and model, the current value depreciation curve calculated based on vehicle age and mileage, and the vehicle's historical maintenance records; The extracted features are cleaned and standardized to convert all features into a structured feature vector. Feature selection and dimensionality reduction are performed on the structured feature vector, and the dataset is divided into training, validation and test sets. Based on the divided dataset, random forest is selected for model training and hyperparameter tuning to learn the mapping relationship between customer features and credit risk. The risk probability value output by the model is mapped to the final customer credit score, and the credit score is transmitted to the risk assessment module through the API interface.

[0008] Preferably, as a preferred embodiment of the AI-based dynamic risk assessment system for auto finance described in this invention, the loan history analysis module assesses a customer's repayment ability and past default records based on their loan history data, and evaluates their repayment risk, specifically including: The customer information collection module obtains detailed loan history records of customers and extracts overdue days distribution, historical overdue number, loan type, repayment frequency, maximum overdue amount of a single loan, and loan settlement status characteristics from the loan history data. Survival analysis algorithms are used to analyze extracted features, predict the probability of future loan defaults, and generate repayment risk assessment results, including low risk, medium risk, and high risk, further including: The extracted raw features are standardized and processed to construct target variables; The Cox proportional hazards model is used to analyze the risk rates of events with different characteristics. The formula for the risk model is as follows: ,in, This represents the default risk rate of a customer at time point t, given the feature vector X. Represents the benchmark risk function. It is the preprocessed customer feature vector. Let be the parameter vector to be estimated. It's the risk ratio; Substitute the customer's feature vector X into the trained Cox model to calculate their risk score. The specific formula is as follows: ,in, It is a risk score; Calculate the individualized survival probability curve for each customer based on their risk score: ,in, It is the baseline survival function. This represents the probability that a customer has not defaulted by any future time point t. The cumulative probability of default at a specific time point T is derived from the survival curve. Risk stratification is performed using predefined thresholds, including the following: When the risk score is below the lower threshold, the customer is classified as a low-risk repayment customer; when the risk score is between the upper and lower thresholds, the customer is classified as a medium-risk repayment customer; when the risk score is above the upper threshold, the customer is classified as a high-risk repayment customer.

[0009] Preferably, as a preferred embodiment of the AI-based dynamic risk assessment system for auto finance described in this invention, the vehicle valuation module estimates the vehicle's value based on its brand, model, and age, combined with market trends, and determines the vehicle's mortgage risk. Specifically, this includes: The system obtains information on the brand, model, age, and mileage of the vehicle the customer intends to mortgage. Based on a regression model using market transaction data of similar vehicles, it estimates the vehicle's current market value. The specific formula is as follows: ,in, It is the estimated current market value of the vehicle. This is the intercept term. Age is the vehicle's age, Mileage is the mileage driven, and E represents characteristics that affect the vehicle's value, including brand, model code, configuration, and region. These are feature coefficients obtained through model training, representing the degree of influence of each feature on the vehicle's value. It is an error term; Combining vehicle valuation results and maintenance records, the risk of using a vehicle as collateral is assessed, generating a vehicle mortgage risk assessment result. Depreciation risk and market liquidity are input as quantitative factors into the risk assessment model. Each risk factor is assigned a different score and weight according to its risk level, and a comprehensive judgment is made to obtain a risk score. ,in, The depreciation risk score is calculated based on the residual value rate. It is a liquidity risk classification. The weights of each factor are determined through model training, based on the comprehensive risk score. Conduct risk classification and generate an assessment report, the details of which are as follows: when The risk assessment result for the vehicle mortgage was low. when The risk assessment result for the vehicle mortgage is medium risk; when The risk assessment result for the vehicle mortgage was classified as high risk.

[0010] Preferably, as a preferred embodiment of the AI-based dynamic risk assessment system for auto finance described in this invention, the risk assessment module integrates the credit score generated by the customer credit assessment module, the repayment risk determined by the loan history analysis module, and the collateral risk determined by the vehicle valuation module. It utilizes AI algorithms to perform deep fusion analysis on the aforementioned multi-dimensional risk factors, calculates the final comprehensive risk level, and provides suggestions on loan approval, loan amount, and interest rate adjustments based on preset risk thresholds. The final result is then sent to the data storage module for recording. Specifically, this includes: It receives credit scores from the customer credit assessment module, repayment risk assessment results from the loan history analysis module, and mortgage risk assessment results from the vehicle valuation module. It converts the repayment risk level and mortgage risk level into ordinal values ​​and combines them with the credit score to construct a joint feature vector. The specific expression is as follows: The gradient boosting tree algorithm is used to perform deep fusion analysis of multi-dimensional risk factors, and the final comprehensive risk score is calculated. ,in, It's a credit score. It is an ordinal code for the repayment risk level: "Low repayment risk = 0, Medium repayment risk = 1, High repayment risk = 2". It is an ordinal code for the mortgage risk level: "Low mortgage risk = 0, medium mortgage risk = 1, high mortgage risk = 2". K represents the current estimated value of the vehicle, and K is the total number of regression trees. It is the predicted output of the k-th tree; It is the function space of the regression tree; Based on preset risk thresholds and business strategies, automated suggestions are provided for loan approval. These suggestions include combining comprehensive risk level, customer repayment ability, and vehicle valuation to provide the maximum loan amount. The corresponding loan interest rate tier is matched according to the risk level to achieve risk pricing. The final decision suggestions are sent to the data storage module for permanent recording and returned to the front-end business system for display and manual review.

[0011] This application also provides an AI-based dynamic risk assessment method for auto finance, the method specifically including: Collect basic customer information, including credit score, income status, loan history, vehicle brand, model, and age. Machine learning algorithms are used to process customer information, assess customer credit risk, and generate corresponding credit scores. Based on the customer's loan history data, assess the customer's repayment ability and past default records to evaluate their repayment risk; Based on the vehicle's brand, model, and year of manufacture, combined with market trends, the vehicle is valued, and the mortgage risk is assessed. By comprehensively considering credit scores, repayment risks, and collateral risks, AI algorithms are used to deeply integrate and analyze these multi-dimensional risk factors to calculate the final comprehensive risk level. Based on preset risk thresholds, suggestions are given on whether to approve the loan, the loan amount, and interest rate adjustments.

[0012] On the other hand, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements the functional modules of an AI-based dynamic risk assessment system for automobile finance as described above.

[0013] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements an AI-based dynamic risk assessment system for automotive finance as described above.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: By using machine learning algorithms to deeply process multi-dimensional information such as customer credit scores, loan history, and income status, it is possible to more accurately assess customer credit risk, reduce the subjectivity and bias in manual assessment, and provide financial institutions with more accurate vehicle valuation and collateral risk assessment by comprehensively analyzing factors such as vehicle brand, model, age, and market trends. This helps determine whether a vehicle is suitable as collateral for a loan and effectively reduces vehicle mortgage risk. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0016] Figure 1 This is a flowchart of the method of the present invention.

[0017] Table 1 is a data recording table of the simulation experiment of this invention. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0020] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0021] Example 1 This embodiment provides, for example Figure 1 The system shown is an AI-based dynamic risk assessment system for auto finance, which includes a customer information collection module, a customer credit assessment module, a loan history analysis module, a vehicle valuation module, a risk assessment module, and a data storage module. The customer information collection module collects basic customer information, including credit score, income status, loan history, vehicle brand, model, and age, and sends the customer information to the customer credit assessment module. The customer credit assessment module uses machine learning algorithms to process the received customer information, assess the customer's credit risk, and generate a corresponding credit score. The loan history analysis module is based on the customer's loan history data to determine the customer's repayment ability and past default records, and to assess their repayment risk. The vehicle valuation module is based on the vehicle's brand, model, and year of manufacture, combined with market trends, to value the vehicle and assess its mortgage risk. The risk assessment module integrates the credit score generated by the customer credit assessment module, the repayment risk judged by the loan history analysis module, and the mortgage risk judged by the vehicle valuation module. It uses AI algorithms to perform in-depth fusion analysis on the above multi-dimensional risk factors, calculates the final comprehensive risk level, and provides suggestions on whether to approve the loan, the loan amount, and interest rate adjustments based on preset risk thresholds. The final result is then sent to the data storage module for recording.

[0022] In this embodiment, the customer information collection module needs to be specifically described. This module collects basic customer information, including credit score, income status, loan history, vehicle brand, model, and age, and sends this information to the customer credit assessment module. Specifically, this includes: The system collects basic identity information, personal financial information, and vehicle information of customers. The basic identity information includes name, ID number, contact information, and address. The personal financial information includes credit report data, income data, and loan history data. The vehicle information includes the brand, model, age, mileage, and VIN code information of the vehicle that the customer intends to purchase or mortgage. The collected data undergoes preliminary formatting and basic data integrity and compliance verification. The pre-processed and verified customer basic information, credit report data, income data, loan history data, and vehicle information are then securely packaged and sent to the customer credit assessment module.

[0023] In this embodiment, the customer credit assessment module needs to be specifically described. This module uses machine learning algorithms to process received customer information, assess the customer's credit risk, and generate a corresponding credit score. Specifically, it includes: It receives basic customer identity information, personal financial information, and vehicle information from the customer information collection module, which have been preliminarily formatted and verified. It extracts key features from the received raw data, including credit report features, income and debt features, loan history features, and vehicle features, in order to maximize information utilization efficiency and adapt to machine learning models. The credit report features include historical credit scores, number of overdue payments and days, number of credit accounts, credit card limit utilization rate, number of loan applications, multiple borrowing, and number of credit inquiries; the income and debt features include average monthly income, income source type, and income-to-debt ratio; the loan history features include total historical loan amount, amount repaid, outstanding balance, loan term, and number of prepayments; the vehicle features include calculating the residual value rate of the vehicle brand and model, the current value depreciation curve calculated based on vehicle age and mileage, and the vehicle's historical maintenance records; The extracted features are cleaned and standardized to convert all features into a structured feature vector. Feature selection and dimensionality reduction are performed on the structured feature vector, and the dataset is divided into training, validation and test sets. Based on the divided dataset, random forest is selected for model training and hyperparameter tuning to learn the mapping relationship between customer features and credit risk. The risk probability value output by the model is mapped to the final customer credit score, and the credit score is transmitted to the risk assessment module through the API interface.

[0024] In this embodiment, the loan history analysis module needs to be specifically described. This module assesses a customer's repayment ability and past default records based on their loan history data, and evaluates their repayment risk. Specifically, it includes: The customer information collection module obtains detailed loan history records of customers and extracts overdue days distribution, historical overdue number, loan type, repayment frequency, maximum overdue amount of a single loan, and loan settlement status characteristics from the loan history data. Survival analysis algorithms are used to analyze extracted features, predict the probability of future loan defaults, and generate repayment risk assessment results, including low risk, medium risk, and high risk, further including: The extracted raw features are standardized and processed to construct target variables; The Cox proportional hazards model is used to analyze the risk rates of events with different characteristics. The formula for the risk model is as follows: ,in, This represents the default risk rate of a customer at time point t, given the feature vector X. Represents the benchmark risk function. It is the preprocessed customer feature vector. Let be the parameter vector to be estimated. It's the risk ratio; Substitute the customer's feature vector X into the trained Cox model to calculate their risk score. The specific formula is as follows: ,in, It is a risk score; Calculate the individualized survival probability curve for each customer based on their risk score: ,in, It is the baseline survival function. This represents the probability that a customer has not defaulted by any future time point t. The cumulative probability of default at a specific time point T is derived from the survival curve. Risk stratification is performed using predefined thresholds, including the following: When the risk score is below the lower threshold, the customer is classified as a low-risk repayment customer; when the risk score is between the upper and lower thresholds, the customer is classified as a medium-risk repayment customer; when the risk score is above the upper threshold, the customer is classified as a high-risk repayment customer.

[0025] In this embodiment, the vehicle valuation module needs to be specifically explained. This module values ​​the vehicle based on its brand, model, and age, combined with market trends, and assesses the vehicle's mortgage risk. Specifically, it includes: The system obtains information on the brand, model, age, and mileage of the vehicle the customer intends to mortgage. Based on a regression model using market transaction data of similar vehicles, it estimates the vehicle's current market value. The specific formula is as follows: ,in, It is the estimated current market value of the vehicle. This is the intercept term. Age is the vehicle's age, Mileage is the mileage driven, and E represents characteristics that affect the vehicle's value, including brand, model code, configuration, and region. These are feature coefficients obtained through model training, representing the degree of influence of each feature on the vehicle's value. It is an error term; Combining vehicle valuation results and maintenance records, the risk of using a vehicle as collateral is assessed, generating a vehicle mortgage risk assessment result. Depreciation risk and market liquidity are input as quantitative factors into the risk assessment model. Each risk factor is assigned a different score and weight according to its risk level, and a comprehensive judgment is made to obtain a risk score. ,in, The depreciation risk score is calculated based on the residual value rate. It is a liquidity risk classification. The weights of each factor are determined through model training, based on the comprehensive risk score. Conduct risk classification and generate an assessment report, the details of which are as follows: when The risk assessment result for the vehicle mortgage was low. when The risk assessment result for the vehicle mortgage is medium risk; when The risk assessment result for the vehicle mortgage was classified as high risk.

[0026] In this embodiment, the risk assessment module is specifically described. This module integrates the credit score generated by the customer credit assessment module, the repayment risk assessed by the loan history analysis module, and the collateral risk assessed by the vehicle valuation module. It uses AI algorithms to perform deep fusion analysis on these multi-dimensional risk factors, calculates the final comprehensive risk level, and provides suggestions on loan approval, loan amount, and interest rate adjustments based on preset risk thresholds. The final result is then sent to the data storage module for recording. Specifically, this includes: It receives credit scores from the customer credit assessment module, repayment risk assessment results from the loan history analysis module, and mortgage risk assessment results from the vehicle valuation module. It converts the repayment risk level and mortgage risk level into ordinal values ​​and combines them with the credit score to construct a joint feature vector. The specific expression is as follows: The gradient boosting tree algorithm is used to perform deep fusion analysis of multi-dimensional risk factors, and the final comprehensive risk score is calculated. ,in, It's a credit score. It is an ordinal code for the repayment risk level: "Low repayment risk = 0, Medium repayment risk = 1, High repayment risk = 2". It is an ordinal code for the mortgage risk level: "Low mortgage risk = 0, medium mortgage risk = 1, high mortgage risk = 2". K represents the current estimated value of the vehicle, and K is the total number of regression trees. It is the predicted output of the k-th tree; It is the function space of the regression tree; Based on preset risk thresholds and business strategies, automated suggestions are provided for loan approval. These suggestions include combining comprehensive risk level, customer repayment ability, and vehicle valuation to provide the maximum loan amount. The corresponding loan interest rate tier is matched according to the risk level to achieve risk pricing. The final decision suggestions are sent to the data storage module for permanent recording and returned to the front-end business system for display and manual review.

[0027] This application also provides an AI-based dynamic risk assessment method for auto finance, the method specifically including: Collect basic customer information, including credit score, income status, loan history, vehicle brand, model, and age. Machine learning algorithms are used to process customer information, assess customer credit risk, and generate corresponding credit scores. Based on the customer's loan history data, assess the customer's repayment ability and past default records to evaluate their repayment risk; Based on the vehicle's brand, model, and year of manufacture, combined with market trends, the vehicle is valued, and the mortgage risk is assessed. By comprehensively considering credit scores, repayment risks, and collateral risks, AI algorithms are used to deeply integrate and analyze these multi-dimensional risk factors to calculate the final comprehensive risk level. Based on preset risk thresholds, suggestions are given on whether to approve the loan, the loan amount, and interest rate adjustments.

[0028] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0029] This embodiment also provides a storage medium storing a computer program. When executed by a processor, the program implements the functional modules of an AI-based dynamic risk assessment system for automotive finance as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0030] Example 2 The following is another embodiment of the present invention, which provides an AI-based dynamic risk assessment system for automotive finance. To verify the beneficial effects of the present invention, a simulation experiment is conducted for scientific demonstration.

[0031] This experiment aims to verify the effectiveness of an AI-based dynamic risk assessment system for auto finance. Through technical means such as customer information collection, credit assessment, loan history analysis, vehicle valuation, risk assessment, and decision analysis, it enhances the risk management capabilities in the auto finance sector. The experiment uses simulated and actual collected customer data, vehicle information, and market change data. By analyzing the consistency between the model's risk prediction results and actual loan decisions, the experiment verifies the system's accuracy and robustness in identifying and predicting financial risks. The simulation experiment steps are implemented according to the content of the AI-based dynamic risk assessment system for auto finance provided in Example 1, and the specific steps include: Collect customer information data, including customer credit score, income status, loan history data, vehicle brand, model, and age information, with the collection frequency set to each loan application. The received customer information is processed using machine learning algorithms to assess the customer's credit risk and generate a corresponding credit score. Based on the customer's loan history data, assess their repayment ability and past default records, and analyze their repayment risk; Based on the vehicle's brand, model, and year of manufacture, combined with market trends, the vehicle is valued, and its mortgage risk is assessed. By integrating data from three modules—customer credit assessment, loan history analysis, and vehicle valuation—AI algorithms are used to perform in-depth analysis of the aforementioned multi-dimensional risk factors to calculate the final comprehensive risk level. Based on preset risk thresholds, suggestions are given on whether to approve the loan, the loan amount, and interest rate adjustments, and the final results are sent to the data storage module for recording.

[0032] The specific data from the above simulation experiment are as follows: Customer credit score Income status Loan history Vehicle Information (Brand / Model / Year) Risk assessment weight Risk assessment results Loan approval or not Loan amount / interest rate Risk management measures 720 high income good BMW 320i 3 years 0.85 Low risk approve 200,000 yuan / 5% Normal approval 650 Middle-income One breach of contract Toyota / Corolla / 5 years 0.78 Medium risk no none Strengthen review 580 low income Multiple breaches of contract Mercedes-Benz C200 / 6 years 0.92 High risk no none Deny loan 740 high income good Audi / A4 / 2 years 0.80 Low risk approve 180,000 yuan / 4.5% Normal approval 710 Middle-income One breach of contract Honda / Accord / 4 years 0.75 Medium risk no none Strengthen review 800 high income No breach of contract Mercedes-Benz / E-Class / 3 years 0.88 Low risk approve 250,000 yuan / 4% Normal approval Table 1 Experimental Analysis: By comparing the consistency between the model's predictions and actual loan decisions, the accuracy and robustness of the system in identifying and predicting auto finance risks are verified. Experiments show that the AI-based dynamic risk assessment system for auto finance can accurately assess customers' credit risk, repayment risk, and collateral risk. By comparing with actual loan decisions, it demonstrates high accuracy and reliability, effectively improving auto finance risk management capabilities. Experimental results indicate that the AI-based dynamic risk assessment system provided by this invention possesses high risk prediction capabilities and decision support levels in the auto finance field, effectively improving the accuracy of loan decisions and the efficiency of financial risk control.

[0033] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. An AI-based dynamic risk assessment system for auto finance, characterized in that: It includes a customer information collection module, a customer credit assessment module, a loan history analysis module, a vehicle valuation module, a risk assessment module, and a data storage module; The customer information collection module collects basic customer information, including credit score, income status, loan history, vehicle brand, model, and age, and sends the customer information to the customer credit assessment module. The customer credit assessment module uses machine learning algorithms to process the received customer information, assess the customer's credit risk, and generate a corresponding credit score. The loan history analysis module is based on the customer's loan history data to determine the customer's repayment ability and past default records, and to assess their repayment risk. The vehicle valuation module is based on the vehicle's brand, model, and year of manufacture, combined with market trends, to value the vehicle and assess its mortgage risk. The risk assessment module integrates the credit score generated by the customer credit assessment module, the repayment risk judged by the loan history analysis module, and the mortgage risk judged by the vehicle valuation module. It uses AI algorithms to perform in-depth fusion analysis on the above multi-dimensional risk factors, calculates the final comprehensive risk level, and provides suggestions on whether to approve the loan, the loan amount, and interest rate adjustments based on preset risk thresholds. The final result is then sent to the data storage module for recording.

2. The AI-based dynamic risk assessment system for auto finance according to claim 1, characterized in that: The customer information collection module collects basic customer information, including credit score, income status, loan history, vehicle brand, model, and age, and sends this information to the customer credit assessment module, specifically including: The system collects basic identity information, personal financial information, and vehicle information of customers. The basic identity information includes name, ID number, contact information, and address. The personal financial information includes credit report data, income data, and loan history data. The vehicle information includes the brand, model, age, mileage, and VIN code information of the vehicle that the customer intends to purchase or mortgage. The collected data undergoes preliminary formatting and basic data integrity and compliance verification. The pre-processed and verified customer basic information, credit report data, income data, loan history data, and vehicle information are then securely packaged and sent to the customer credit assessment module.

3. The AI-based dynamic risk assessment system for auto finance according to claim 1, characterized in that: The customer credit assessment module uses machine learning algorithms to process received customer information, assess customer credit risk, and generate corresponding credit scores, specifically including: It receives basic customer identity information, personal financial information, and vehicle information from the customer information collection module, which have been preliminarily formatted and verified. It extracts key features from the received raw data, including credit report features, income and debt features, loan history features, and vehicle features, in order to maximize information utilization efficiency and adapt to machine learning models. The extracted features are cleaned and standardized to convert all features into a structured feature vector. Feature selection and dimensionality reduction are performed on the structured feature vector, and the dataset is divided into training, validation and test sets. Based on the divided dataset, random forest is selected for model training and hyperparameter tuning to learn the mapping relationship between customer features and credit risk. The risk probability value output by the model is mapped to the final customer credit score, and the credit score is transmitted to the risk assessment module through the API interface.

4. The AI-based dynamic risk assessment system for auto finance according to claim 1, characterized in that: The loan history analysis module assesses a customer's repayment ability and past default records based on their loan history data, and evaluates their repayment risk. Specifically, it includes: The customer information collection module obtains detailed loan history records of customers and extracts overdue days distribution, historical overdue number, loan type, repayment frequency, maximum overdue amount of a single loan, and loan settlement status characteristics from the loan history data. Survival analysis algorithms are used to analyze extracted features, predict the probability of future loan defaults, and generate repayment risk assessment results, including low risk, medium risk, and high risk, further including: The extracted raw features are standardized and processed to construct target variables; The Cox proportional hazards model is used to analyze the risk rates of events with different characteristics. The formula for the risk model is as follows: ,in, This represents the default risk rate of a customer at time point t, given the feature vector X. Represents the benchmark risk function. It is the preprocessed customer feature vector. Let be the parameter vector to be estimated. It's the risk ratio; Substitute the customer's feature vector X into the trained Cox model to calculate their risk score. The specific formula is as follows: ,in, It is a risk score; Calculate the individualized survival probability curve for each customer based on their risk score: ,in, It is the baseline survival function. This represents the probability that a customer has not defaulted by any future time point t. The cumulative probability of default at a specific time point T is derived from the survival curve. Risk stratification is performed using predefined thresholds, including the following: When the risk score is below the lower threshold, the customer is classified as a low-risk repayment customer; when the risk score is between the upper and lower thresholds, the customer is classified as a medium-risk repayment customer; when the risk score is above the upper threshold, the customer is classified as a high-risk repayment customer.

5. The AI-based dynamic risk assessment system for auto finance according to claim 1, characterized in that: The vehicle valuation module values ​​the vehicle based on its brand, model, and age, combined with market trends, and assesses the vehicle's mortgage risk. Specifically, it includes: The system obtains information on the brand, model, age, and mileage of the vehicle the customer intends to mortgage. Based on a regression model using market transaction data of similar vehicles, it estimates the vehicle's current market value. The specific formula is as follows: ,in, It is the estimated current market value of the vehicle. This is the intercept term. Age is the vehicle's age, Mileage is the mileage driven, and E represents characteristics that affect the vehicle's value, including brand, model code, configuration, and region. These are feature coefficients obtained through model training, representing the degree of influence of each feature on the vehicle's value. It is an error term; Combining vehicle valuation results and maintenance records, the risk of using a vehicle as collateral is assessed, generating a vehicle mortgage risk assessment result. Depreciation risk and market liquidity are input as quantitative factors into the risk assessment model. Each risk factor is assigned a different score and weight according to its risk level, and a comprehensive judgment is made to obtain a risk score. ,in, The depreciation risk score is calculated based on the residual value rate. It is a liquidity risk classification. The weights of each factor are determined through model training, based on the comprehensive risk score. Conduct risk classification and generate an assessment report, the details of which are as follows: when The risk assessment result for the vehicle mortgage was low. when The risk assessment result for the vehicle mortgage is medium risk; when The risk assessment result for the vehicle mortgage was classified as high risk.

6. The AI-based dynamic risk assessment system for auto finance according to claim 1, characterized in that: The risk assessment module integrates the credit score generated by the customer credit assessment module, the repayment risk assessed by the loan history analysis module, and the collateral risk assessed by the vehicle valuation module. It uses AI algorithms to perform deep fusion analysis of these multi-dimensional risk factors to calculate the final comprehensive risk level. Based on preset risk thresholds, it provides suggestions on loan approval, loan amount, and interest rate adjustments, and sends the final result to the data storage module for recording. Specifically, this includes: It receives credit scores from the customer credit assessment module, repayment risk assessment results from the loan history analysis module, and mortgage risk assessment results from the vehicle valuation module. It converts the repayment risk level and mortgage risk level into ordinal values ​​and constructs a joint feature vector with the credit score. It then uses the gradient boosting tree algorithm to perform deep fusion analysis of multi-dimensional risk factors and calculates the final comprehensive risk score. Based on preset risk thresholds and business strategies, automated suggestions are provided for loan approval. These suggestions include combining comprehensive risk level, customer repayment ability, and vehicle valuation to provide the maximum loan amount. The corresponding loan interest rate tier is matched according to the risk level to achieve risk pricing. The final decision suggestions are sent to the data storage module for permanent recording and returned to the front-end business system for display and manual review.

7. The AI-based dynamic risk assessment system for auto finance according to claim 6, characterized in that: The joint credit score is constructed as a joint feature vector, specifically expressed as follows: The gradient boosting tree algorithm is used to perform deep fusion analysis of multi-dimensional risk factors, and the final comprehensive risk score is calculated. ,in, It's a credit score. It is an ordinal code for the repayment risk level: "Low repayment risk = 0, Medium repayment risk = 1, High repayment risk = 2". It is an ordinal code for the mortgage risk level, "low risk = 0, medium risk = 1, high risk = 2". K represents the current estimated value of the vehicle, and K is the total number of regression trees. It is the predicted output of the k-th tree; It is the function space of the regression tree.

8. An AI-based dynamic risk assessment method for auto finance is applied to an AI-based dynamic risk assessment system for auto finance as described in any one of claims 1-7, characterized in that: Specifically, it includes: Collect basic customer information, including credit score, income status, loan history, vehicle brand, model, and age. Machine learning algorithms are used to process customer information, assess customer credit risk, and generate corresponding credit scores. Based on the customer's loan history data, assess the customer's repayment ability and past default records to evaluate their repayment risk; Based on the vehicle's brand, model, and year of manufacture, combined with market trends, the vehicle is valued, and the mortgage risk is assessed. By comprehensively considering credit scores, repayment risks, and collateral risks, AI algorithms are used to deeply integrate and analyze these multi-dimensional risk factors to calculate the final comprehensive risk level. Based on preset risk thresholds, suggestions are given on whether to approve the loan, the loan amount, and interest rate adjustments.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the functional modules of the AI-based dynamic risk assessment system for automotive finance as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the functional modules of the AI-based dynamic risk assessment system for automotive finance as described in any one of claims 1-7.

Images