Insurance control value analysis method, apparatus and device for insurance applicant, and medium

Abstract

The invention relates to the technical field of intelligent decision making, can be applied to business system platforms of financial science and technology, medical health and the like, and discloses an insurance control value analysis method, device, equipment and medium for an insurance applicant, and the method comprises the following steps: obtaining multi-dimensional risk data of the insurance applicant, and carrying out implicit association feature cross extraction on the multi-dimensional risk data to obtain feature association data; determining a comprehensive risk index of the multi-dimensional risk data according to the feature association data, and performing correlation analysis on the comprehensive risk index and the multi-dimensional risk data to obtain an analysis result; performing data extraction on the multi-dimensional risk data according to an analysis result to obtain key risk data; and constructing a condition set according to the key risk data, and calculating an initial control value of the multi-dimensional risk data according to the condition set and the key risk data. Through hidden correlation feature cross extraction, a richer basis is provided for risk assessment and decision making, and meanwhile, a condition set is adjusted and optimized, so that a risk management system keeps dynamic adaptability.

Description

Technical Field

[0001] The present invention relates to the field of intelligent decision-making technology, and in particular to a method, device, equipment and medium for analyzing the insurance control value of an insured. Background Art

[0002] Policyholder insurance refers to a commercial insurance behavior in which the policyholder pays premiums to the insurer in accordance with the contract, and the insurer bears the responsibility of paying insurance compensation for property losses caused by accidents that may occur as stipulated in the contract; policyholder insurance includes medical insurance and property insurance. Policyholder medical insurance is an insurance product purchased by the policyholder, which is an insurance product that protects the medical expenses of the insured. Policyholder property insurance is an insurance product in which the policyholder takes the property and related interests as the insured object, pays premiums to the insurer, and the insurer bears the responsibility of paying insurance compensation for property losses caused by accidents that may occur as stipulated in the contract.

[0003] At present, during the policyholder acceptance process, whether it is property insurance for policyholders oriented towards the financial field or medical insurance for policyholders oriented towards the medical and health field, insurance institutions generally conduct risk analysis based on machine learning ideas and formulate an insurance premium. However, when formulating policyholder insurance premiums, the existing method only combines the policyholder's single data dimension with fixed calculation rules to price the insurance premium. Although this pricing model is simple, it cannot fully capture various risk factors for the policyholder's complex risk data, nor can it adjust the rules based on the policyholder's risk data, resulting in a vague judgment on the policyholder's insurance premium.

[0004] Therefore, in the face of the growing market demand for policyholder insurance, the current policyholder insurance control value analysis method urgently needs to be improved to solve the problem that the existing method is difficult to deeply analyze the risk characteristics implicit in the policyholder risk data and the analysis results are not accurate enough. Summary of the Invention

[0005] The present invention provides a method, device, equipment and medium for analyzing the insurance control value of an insured. Through cross-extraction of implicit correlation features, it provides a richer basis for risk assessment and decision-making, and at the same time adjusts and optimizes the condition set to enable the risk management system to maintain dynamic adaptability.

[0006] In a first aspect, a method for analyzing an insurance control value of an insured is provided, comprising:

[0007] Acquire multidimensional risk data of the insured, and perform cross-extraction of implicit correlation features on the multidimensional risk data to obtain feature correlation data;

[0008] Determining a comprehensive risk index of the multidimensional risk data based on the characteristic association data, and performing a correlation analysis between the comprehensive risk index and the multidimensional risk data to obtain an analysis result;

[0009] Extracting the multidimensional risk data according to the analysis results to obtain key risk data;

[0010] Constructing a condition set according to the key risk data, and calculating an initial control value of the multi-dimensional risk data according to the condition set and the key risk data;

[0011] Optimizing condition parameters of the condition set according to the initial control value to obtain an adjusted condition set;

[0012] A target control value of the multi-dimensional risk data is calculated according to the adjustment condition set and the key risk data.

[0013] In a second aspect, a device for analyzing an insurance control value of an insured is provided, comprising:

[0014] Acquisition module, used to obtain the policyholder's multi-dimensional risk data;

[0015] An extraction module, configured to perform cross-extraction of implicit correlation features on the multi-dimensional risk data to obtain feature correlation data;

[0016] An indicator determination module, configured to determine a comprehensive risk indicator of the multi-dimensional risk data based on the feature association data;

[0017] An analysis module, configured to perform a correlation analysis on the comprehensive risk indicator and the multi-dimensional risk data to obtain an analysis result;

[0018] A data extraction module, configured to extract the multidimensional risk data according to the analysis results to obtain key risk data;

[0019] A construction module, configured to construct a condition set based on the key risk data;

[0020] an initial calculation module, configured to calculate an initial control value of the multi-dimensional risk data according to the condition set and the key risk data;

[0021] The optimization module is configured to optimize the condition parameters of the condition set according to the initial control value to obtain an adjusted condition set.

[0022] A target calculation module calculates a target control value of the multi-dimensional risk data according to the adjustment condition set and the key risk data.

[0023] In a third aspect, a computer device is provided, comprising 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 method for analyzing the insurance control value of an insured when executing the computer program.

[0024] In a fourth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned method for analyzing the insurance control value of an insured are implemented.

[0025] In the scheme implemented by the above-mentioned insured insurance control value analysis method, device, computer equipment and storage medium, there are various explicit and implicit correlations in the multidimensional risk data. Explicit correlations are easy to be discovered, but implicit correlations are often hidden deep in the data. Through cross-extraction of implicit correlation features, these subtle relationships can be excavated, the potential laws and patterns in the data can be discovered, and more comprehensive and in-depth information can be obtained to provide a richer basis for risk assessment and decision-making. At the same time, by analyzing the correlation between the comprehensive risk indicators and the various dimensions of multidimensional risk data, the contribution degree and mutual relationship of different risk factors to the overall risk can be clearly grasped, thereby formulating differentiated risk management strategies for different risk dimensions. In addition, by optimizing the calculation rule set based on the initial control value, the risk assessment model can be adjusted in time according to new risk data and business conditions, so that the risk management system maintains dynamic adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0027] Figure 1 This is a schematic diagram of an application environment of a method for analyzing insurance control value of an insured in one embodiment of the present invention;

[0028] Figure 2 This is a flow chart of a method for analyzing an insurance control value of an insured according to an embodiment of the present invention;

[0029] Figure 3 This is a structural diagram of a device for analyzing insurance control value of an insured according to an embodiment of the present invention;

[0030] Figure 4 is a structural diagram of a computer device according to an embodiment of the present invention;

[0031] Figure 5 It is another structural schematic diagram of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0032] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0033] The embodiment of the present invention provides a method for analyzing the insurance control value of an insured person, which can be applied in the following cases: Figure 1 In an application environment, the client communicates with the server through a network. The server can obtain the insured's multidimensional risk data, perform implicit correlation feature cross-extraction on the multidimensional risk data, and obtain feature correlation data; determine the comprehensive risk index of the multidimensional risk data based on the feature correlation data, and perform correlation analysis on the comprehensive risk index and the multidimensional risk data to obtain analysis results; extract data from the multidimensional risk data based on the analysis results to obtain key risk data; construct a condition set based on the key risk data, and calculate the initial control value of the multidimensional risk data based on the condition set and the key risk data; optimize the condition parameters of the condition set based on the initial control value to obtain an adjusted condition set; calculate the target control value of the multidimensional risk data based on the adjusted condition set and the key risk data, and feed the target control value back to the client. The present invention provides an insured insurance control value analysis device. For target control value business, the device provides a richer basis for risk assessment and decision-making through implicit correlation feature cross-extraction, and at the same time adjusts and optimizes the condition set to keep the risk management system dynamic and adaptable. The client can be, but is not limited to, various personal computers, laptops, smartphones, tablet computers, and portable wearable devices. The server can be implemented as an independent server or a server cluster consisting of multiple servers. The present invention is described in detail below through specific embodiments.

[0034] See also Figure 2 As shown, Figure 2 A flowchart of a method for analyzing an insurance control value of an insured provided by an embodiment of the present invention includes the following steps:

[0035] S1. Obtain multi-dimensional risk data of the insured, perform cross-extraction of implicit correlation features on the multi-dimensional risk data, and obtain feature correlation data.

[0036] In an embodiment of the present invention, the acquisition refers to collecting data information of multiple dimensions related to the risk status of the insured, and the cross-extraction of hidden correlation features refers to the process of extracting the hidden and difficult to directly observe correlation relationships in the multi-dimensional risk data in the form of features.

[0037] Specifically, multidimensional risk data includes the insured's basic information, the insured's insurance label information, health status data, and behavioral habit data. Specific data mining and analysis methods are used to discover the hidden and non-intuitive relationships in these multidimensional risk data, and extract these relationships in the form of features. The final result is feature-related data, which can present the inherent structure and potential laws of the insured's risks in a more in-depth and comprehensive manner.

[0038] In specific medical and health scenarios, we obtain multi-dimensional risk data on the insured, including age, gender, race, family medical history, lifestyle habits (such as smoking, drinking, exercise frequency, and eating habits). By analyzing a large amount of patient data, we can discover the implicit correlation between certain lifestyle habits and specific diseases. By conducting a comprehensive analysis of these data, we can predict the probability of an individual developing a certain disease.

[0039] In the fintech scenario, by analyzing consumers' multi-dimensional data, we can discover the implicit correlation between income stability, debt level and credit default. By extracting these characteristic correlation data, financial institutions can more accurately assess customers' credit risks and formulate corresponding credit policies.

[0040] In an embodiment of the present invention, the step of performing implicit correlation feature cross-extraction on the multi-dimensional risk data to obtain feature correlation data includes:

[0041] Performing a rough risk feature screening on the multi-dimensional risk data to obtain retained feature data;

[0042] Calculating a correlation coefficient of the retained feature data, and performing redundant feature recognition on the retained feature data according to the correlation coefficient to obtain recognition feature data;

[0043] Performing high-order cross-combination on the identification feature data to obtain combined data;

[0044] Performing correlation feature extraction on the combined data to obtain feature correlation data.

[0045] In an embodiment of the present invention, the rough screening of risk features refers to the preliminary screening of features related to risk assessment from multidimensional risk data, the calculation refers to the process of calculating the degree of correlation between each retained feature data, the redundant feature identification refers to the process of identifying and eliminating those features in the retained feature data that are highly correlated with other features, have overlapping information content, and do not contribute much to the model or analysis; the high-order cross-recognition refers to the process of multi-level and multi-dimensional combination of different features in the identified feature data, and the associated feature extraction refers to the process of mining features that are closely related to the target variable or other related factors from the combined data.

[0046] Specifically, the data is first cleaned and normalized. Cleaning includes removing noise, missing values and outliers in multidimensional risk data; for example, some obviously unreasonable data points caused by data collection errors or equipment failures are corrected or deleted; for missing values, appropriate filling methods can be selected according to the specific situation, such as mean filling, median filling or filling based on similar samples, etc. Normalization refers to unifying data of different dimensions into the same scale range to eliminate differences in data dimensions and value ranges.

[0047] In detail, based on business knowledge and understanding of risks, features related to risk assessment are preliminarily screened out. For example, in the insurance business, the insured's age, health status, occupation, insurance history, etc. may be features related to insurance risks. Subsequently, by calculating the correlation coefficient between features and other methods, redundant features that are highly correlated with other features are identified and removed to reduce the dimension and amount of calculation of the data, while avoiding the interference of repeated information between features on the analysis results.

[0048] Furthermore, higher-order feature crossovers, such as three-feature combinations and four-feature combinations, can be performed. However, as the order increases, computational complexity and data sparsity become increasingly severe. At the same time, selected features are combined in pairs to form new feature combinations. This approach allows for trade-offs based on actual conditions, using various data mining algorithms, such as association rule mining, cluster analysis, principal component analysis, and factor analysis, to discover hidden associations within feature crossovers. Taking association rule mining as an example, by setting thresholds such as support and confidence, frequent feature association patterns with a certain degree of confidence are identified. Machine learning models, such as decision trees, random forests, and neural networks, are then used to automatically learn the complex relationships between features. During training, the model identifies which feature combinations have a significant impact on the target variable (such as risk assessment results), thereby indirectly mining hidden association features and ultimately generating feature association data.

[0049] In an embodiment of the present invention, through rough screening, those obviously irrelevant or redundant features can be removed, the dimension and complexity of the data can be reduced, and subsequent analysis and processing can be made more efficient. At the same time, identifying and removing redundant features can avoid the reuse of similar information in the model, thereby reducing the risk of overfitting of the model and improving the generalization ability of the model.

[0050] In the embodiments of the present invention, multidimensional risk data contains various explicit and implicit relationships. Explicit relationships are easy to detect, but implicit relationships are often hidden deep within the data. Cross-extraction of implicit relationship features can uncover these subtle relationships, discover underlying patterns and patterns within the data, and obtain more comprehensive and in-depth information, providing a richer basis for risk assessment and decision-making.

[0051] S2. Determine a comprehensive risk index of the multidimensional risk data based on the characteristic association data, and perform a correlation analysis between the comprehensive risk index and the multidimensional risk data to obtain an analysis result.

[0052] In an embodiment of the present invention, the determination refers to the use of a specific mathematical method or model to process and calculate the feature-related data to obtain a numerical value or indicator that can comprehensively reflect the overall risk status of the multi-dimensional risk data, and the correlation analysis refers to determining the relationship between the comprehensive risk indicator and the risk data of each dimension.

[0053] Specifically, based on the various risk feature information contained in the feature-related data and their interrelationships, appropriate algorithms, formulas or models are selected, and these data are integrated, weighted, calculated, and processed to ultimately obtain an indicator that can represent the comprehensive risk level of the multidimensional risk data. Subsequently, statistics such as the correlation coefficient are calculated to measure the closeness of the linear relationship between the comprehensive risk indicator and each variable in the multidimensional risk data.

[0054] In the specific scenario of healthcare, comprehensive risk indicators can be used to assess an individual's risk of developing a specific disease. Based on this indicator, early intervention can be carried out for high-risk patients, such as developing personalized diet and exercise plans and conducting regular targeted examinations to prevent the occurrence of the disease or delay the progression of the disease.

[0055] At the same time, in the investment field, comprehensive risk indicators can be used to assess the risk level of investment projects or investment portfolios. By calculating comprehensive risk indicators, fund managers can understand the potential risks of investment projects in different market environments, thereby adjusting the configuration of investment portfolios, reducing the proportion of higher-risk assets, and improving the stability and return level of the investment portfolio.

[0056] In an embodiment of the present invention, determining the comprehensive risk index of the multi-dimensional risk data based on the feature association data includes:

[0057] Evaluate the importance of each feature-associated data in multidimensional risk data;

[0058] According to the importance of the feature-related data, weight configuration is performed on the feature-related data one by one to obtain the weight corresponding to the feature-related data;

[0059] The weight is weighted and summed with the feature-related data to obtain a comprehensive risk index.

[0060] In an embodiment of the present invention, the weight configuration refers to the process of assigning corresponding numerical weights to each feature-related data based on its relative importance to the comprehensive risk index, and the weighted summation refers to multiplying each feature-related data by its corresponding weight, and then adding these products to obtain a sum, which is the comprehensive risk index.

[0061] Specifically, the feature association data is standardized, and then the information entropy of each feature association data is calculated using the entropy weight method. Then, its weight is calculated based on the information entropy. The specific formula is as follows:

[0062]

[0063] Among them, e j represents the information entropy of the jth feature-related data, n represents the number of feature-related data, W j Represents the weight of the j-th feature associated data.

[0064] In addition, a combined weighting method can also be used, which combines subjective weighting and objective weighting, and adds the subjective weight and the objective weight according to a certain ratio to obtain the final weight value.

[0065] Specifically, ensure that the feature-related data and corresponding weights have been accurately determined, and that the data format and range are suitable for calculations. For example, the feature-related data may need to be standardized or normalized to avoid the impact of different dimensions on the calculation results. In the field of healthcare, different physiological indicators may have different units and orders of magnitude. For example, the unit of blood pressure is mmHg, while the unit of blood sugar is mmol / L. These data need to be standardized so that they are in the same numerical range for weighted summation calculation. The calculation formula is as follows:

[0066] I=w1x1+w2x2+···+w n x n

[0067] Among them, x1, x2, ···x n Represents feature association data, w1,w2,···w n represents the corresponding weight, and I represents the comprehensive risk index.

[0068] In an embodiment of the present invention, performing a correlation analysis on the comprehensive risk indicator and the multi-dimensional risk data to obtain an analysis result includes:

[0069] Calculating the Pearson correlation coefficient between the comprehensive risk indicator and each dimension data in the multidimensional risk data to obtain a coefficient result;

[0070] A correlation analysis is performed on the comprehensive risk indicator and the multidimensional risk data according to the coefficient result to obtain an analysis result.

[0071] Specifically, assuming there are n samples, for multidimensional risk data, let it contain m dimensions, representing X1, X2, ···X m , each dimension has n observations, that is, X i1 ,X i2 ,···X in (i=1,2,···m), the comprehensive risk index is marked as Y, and there are n observations Y1,Y2,···Y n , calculate the mean of each dimension and the mean of the comprehensive risk index respectively, for each dimension X i With the comprehensive risk indicator Y, calculate each dimension X i The covariance of X is calculated at the same time. i The standard deviation of , the standard deviation of Y, and the Pearson correlation coefficient of each dimension data and the comprehensive risk index are calculated according to the formula. The formula is as follows:

[0072]

[0073] Among them, Cov(X i ,Y) represents each dimension X i The covariance of Sx i Represents X i Sy represents the standard deviation of Y, and m coefficient results are obtained.

[0074] Furthermore, the strength of the correlation is judged according to the absolute value of the Pearson correlation coefficient: |r| ≥ 0.8 indicates a strong correlation, 0.5 ≤ |r| < 0.8 indicates a moderate correlation, 0.3 ≤ |r| < 0.5 indicates a weak correlation, and |r| < 0.3 indicates a very weak correlation or almost no linear correlation.

[0075] Specifically, the direction of correlation is determined according to the positive or negative correlation coefficient. If r i >0 means X i Positively correlated with Y, that is, X i When the value of increases, the value of Y also tends to increase; if r i <0, indicating X i Negatively correlated with Y, that is, X i As the value of increases, the value of Y tends to decrease.

[0076] Specifically, the relationship between multidimensional risk data and the comprehensive risk index is comprehensively analyzed by comprehensively considering the strength and direction of the correlation between all dimensional data and the comprehensive risk index. This allows identification of dimensions that have a significant impact on the comprehensive risk index, as well as potential synergistic or antagonistic interactions between dimensions. For example, the finding that dimensions such as age and blood pressure are positively and strongly correlated with the comprehensive disease risk index, while exercise time is negatively correlated with the comprehensive disease risk index, can help doctors or researchers gain a deeper understanding of the factors influencing disease risk and provide a basis for developing prevention and treatment strategies.

[0077] In an embodiment of the present invention, the Pearson correlation coefficient can quantify the relationship between the comprehensive risk indicator and each dimension of the multidimensional risk data, accurately expressing the degree and direction of the linear correlation between them with a numerical value. At the same time, the Pearson correlation coefficient standardizes the data, eliminating the influence of different variable dimensions and value ranges.

[0078] In the embodiment of the present invention, by analyzing the correlation between the comprehensive risk index and each dimension of the multidimensional risk data, the contribution degree and mutual relationship of different risk factors to the overall risk can be clearly grasped, thereby formulating differentiated risk management strategies for different risk dimensions.

[0079] S3. Extract the multidimensional risk data based on the analysis results to obtain key risk data.

[0080] In the embodiment of the present invention, the data extraction refers to the process of selecting a specific data subset that has a significant correlation with the comprehensive risk indicator, ie, key risk data, from the multi-dimensional risk data.

[0081] Specifically, a threshold is set based on the correlation coefficient obtained from the analysis results. For example, the dimensional data corresponding to the correlation coefficient with an absolute value greater than 0.6 is defined as key risk data. Then, the dimensional data that meets the threshold conditions are extracted from the multidimensional risk data. The dimensions that have a greater impact on the comprehensive risk indicators are determined through the analysis results. These dimensions may be multiple. For example, when assessing the operational risk of an enterprise, it is found that the debt-to-asset ratio and current ratio in the financial indicators and the market share in the market indicators are significantly correlated with the comprehensive risk indicators. In this case, the data of these dimensions are extracted as key risk data, and the extracted key risk data are integrated and summarized according to certain rules to facilitate subsequent analysis and application.

[0082] In the specific healthcare context, when assessing the risk of a chronic disease (such as diabetes or cardiovascular disease), multidimensional risk data may include the patient's age, family medical history, lifestyle habits (such as diet, exercise, and smoking), and physical indicators (such as blood pressure, blood sugar, and blood lipids). Correlation analysis may reveal a high correlation between family medical history and blood sugar levels and the risk of diabetes. These two dimensions of data can then serve as key risk data, allowing doctors to provide more targeted disease screening and prevention recommendations for patients.

[0083] In financial scenarios, when assessing the credit risk of individuals or businesses, multidimensional risk data includes income level, debt status, credit history, industry outlook, and corporate financial indicators. Correlation analysis reveals a strong correlation between debt-to-income ratios, credit history, and comprehensive credit risk indicators. These data constitute key risk data. Financial institutions use this key data to decide whether to approve loan applications, as well as the loan amount and interest rate.

[0084] S4. Construct a condition set according to the key risk data, and calculate an initial control value of the multi-dimensional risk data according to the condition set and the key risk data.

[0085] In an embodiment of the present invention, the construction refers to determining a series of calculation rules for describing and defining risk conditions based on the characteristics and analysis requirements of key risk data, and the calculation refers to calculating a numerical value for measuring the overall risk level of multidimensional risk data through a certain algorithm or model based on the constructed condition set and key risk data.

[0086] Specifically, in this solution, the condition set refers to the calculation rule set, and the control value refers to the insurance premium. The insurance institution will study the relationship between the various dimensions of key risk data and the probability of insurance accidents, the degree of loss, etc. Based on the above relationship, the insurance institution will formulate detailed calculation rules. These rules may include different insurance premium adjustment coefficients, risk factor weights, thresholds corresponding to the value ranges of different key risk data, or set specific calculation formulas to comprehensively consider the impact of multiple key risk data on insurance premiums.

[0087] In the specific scenario of medical health, the policy can be formulated based on the relationship between key risk data and the probability of disease occurrence and treatment costs; for example, for insured persons with diabetes, the premium can be adjusted based on factors such as their blood sugar control and whether there are complications.

[0088] In FinTech scenarios, especially in insurance, key risk data varies by insurance type. For property insurance, for example, key risk data might include the value of the insured asset, its useful life, crime rates in the region, and natural disaster rates. Premiums are calculated based on specific risk combinations formed from these key risk data, combined with a set of calculation rules.

[0089] In an embodiment of the present invention, constructing a condition set based on the key risk data includes:

[0090] Objectively weighting the key risk data according to the statistical characteristics of the data to obtain the weight corresponding to the key risk data;

[0091] Performing linear regression analysis on the key risk data to obtain coefficients corresponding to the key risk data;

[0092] Performing statistical distribution analysis on the key risk data, and determining thresholds corresponding to the key risk data based on the statistical distribution analysis results;

[0093] A condition set is generated based on the weights, coefficients and thresholds.

[0094] In an embodiment of the present invention, the objective weighting refers to the process of determining the weights of various indicators of key risk data through specific mathematical methods, the linear regression analysis refers to the process of establishing a linear regression model and finding a set of coefficients so that the model can best fit the data, thereby revealing the degree of influence of key risk data on target variables, the statistical distribution analysis refers to the process of studying the distribution characteristics of key risk data, and the integrated generation refers to the comprehensive processing of information such as the weights obtained through objective weighting, the coefficients obtained through linear regression analysis, and the thresholds determined by statistical distribution analysis to form a unified set of conditions.

[0095] Specifically, the key risk data is standardized and data of different dimensions are converted into comparable values. Assuming that the key risk data has multiple sample sizes and multiple indicator sizes, the entropy weight method is used to calculate the proportion of the i-th sample under the j-th indicator, calculate the entropy value of the j-th indicator, calculate the difference coefficient of the j-th indicator, and finally calculate the weight of the j-th indicator based on the above parameter values.

[0096] In detail, we need to identify the dependent variable and the independent variable. In this scenario, the dependent variable is the premium, and the independent variables are various factors related to the premium, such as the insured’s age, gender, insurance amount, insurance period, etc. These independent variables constitute the various dimensions of the key risk data. Assume that the premium (y) is related to the independent variables (x1, x2, ···x n ) has a linear relationship, which is generally in the form of y = β0 + β1x1 + β2x2 + ··· + βn x n +ε, where β0 is the intercept, β1, β2, ···β n is the regression coefficient, ε is the error term, which represents the difference between the observed value and the true value. The regression coefficient is estimated using methods such as the least squares method, and the regression coefficient is used as the coefficient.

[0097] Furthermore, the mean and standard deviation of the key risk data are calculated, and the mean plus k times the standard deviation is used as the threshold. The weight, coefficient and threshold of each key risk data are clarified. For example, there are three key risk data x1, x2, and x3, and the corresponding weights are w1, w2, and w3, respectively. The coefficients are β1, β2, and β3, respectively. The thresholds are t1, t2, and t3, respectively. Then, conditional rules are formulated based on the business logic and risk assessment model; for example, if w1x1+w2x2+w3x3>T, and x1>t1, x2<t 2. At the same time, if β1x1+β2x2+β3x3>C (C is a value determined according to the linear regression model and business requirements), the situation is classified as a high-risk category, and the corresponding premium adjustment coefficient is a. If the above conditions are not met, but some other conditions are met, it is classified as a medium-risk category or a low-risk category, corresponding to different premium adjustment coefficients b, c, etc. In this way, information such as weights, coefficients, and thresholds are integrated to form a complete calculation rule for evaluating multidimensional risk data and calculating premiums.

[0098] In an embodiment of the present invention, the calculating the initial control value of the multi-dimensional risk data according to the condition set and the key risk data includes:

[0099] Calculating a risk assessment value of the key risk data according to the condition set;

[0100] determining a fuzzy initial control value according to the risk assessment value;

[0101] Linearly combining the weights and coefficients in the condition set to obtain a combination coefficient;

[0102] An initial control value of the multidimensional risk data is calculated according to the fuzzy initial control value and the combination coefficient.

[0103] In the embodiment of the present invention, the linear combination refers to a process of multiplying weights and coefficients.

[0104] Specifically, assuming that the condition set contains key risk data x1, x2, ···x n , the corresponding weights are w1, w2, ···w n , the coefficients are β1, β2, ···β n , using the weighted summation method to calculate the risk assessment value R, the formula is as follows:

[0105]

[0106] Among them, x i * Indicates the result after standardization of key risk data.

[0107] Furthermore, it is necessary to first determine the level of fuzzy control. For example, it can be set to three levels: "high risk", "medium risk" and "low risk". Combined with the threshold determined by historical experience, a corresponding risk assessment value interval is set for each level. According to the calculated risk assessment value, the interval to which it belongs is judged, thereby determining the corresponding fuzzy control level, and assigning an initial control value range to each fuzzy control level. After determining the fuzzy control level, a specific fuzzy initial control value can be randomly selected within the corresponding range or determined according to other rules.

[0108] Furthermore, the weight factor set by the product term of each weight and coefficient is combined with the coefficient vector and the weight vector, and the three are multiplied together. The weighted average method is used to obtain the combined coefficient. Finally, the initial control value of the multidimensional risk data is calculated by multiplying the fuzzy initial control value and the combined coefficient. The final result is the initial control value, that is, the initial premium.

[0109] In this embodiment of the present invention, the weights in the condition set reflect the relative importance of key risk data, while the coefficients embody the quantitative relationship between data and risk. By integrating this information into calculations, risk can be assessed comprehensively and objectively, avoiding assessment bias caused by considering only a single factor. A linear combination of these two key pieces of information, weights and coefficients, generates a single coefficient that comprehensively reflects the impact of key risk data on risk. This consolidates information from multiple dimensions into a concise indicator, facilitating subsequent calculations and analysis while preserving the important information contained in both the weights and coefficients.

[0110] In this embodiment of the present invention, the calculated initial control value is a quantitative indicator that provides a clear reference for risk control. Decision makers can use this value to formulate specific risk control strategies, such as determining the level of resource investment and the risk response measures to be taken. This makes risk control more scientific and precise, improving the efficiency and effectiveness of risk management.

[0111] S5. Optimize condition parameters of the condition set according to the initial control value to obtain an adjusted condition set.

[0112] In the embodiment of the present invention, the condition parameter optimization refers to adjusting and improving various parameters involved in the condition set so that the condition set can better adapt to the actual situation and improve the accuracy and effectiveness of multi-dimensional risk data assessment and control.

[0113] Specifically, the initial control value is calculated based on the original condition set and key risk data. It reflects the initial control level of the current model on multidimensional risk data. However, this initial value may not fully conform to the actual situation or there may be a certain amount of optimization space, and the condition parameters in the condition set need to be optimized.

[0114] In the embodiment of the present invention, the step of optimizing the condition parameters of the condition set according to the initial control value to obtain the adjusted condition set includes:

[0115] Extracting historical claims data from the multidimensional risk data, and performing comparative analysis on the historical claims data and the initial control value to obtain comparative analysis results;

[0116] Performing initial adjustment on condition parameters of the condition set according to the comparative analysis result to obtain an initial adjusted condition set;

[0117] Performing simulation calculation on the initial adjustment condition set and the initial control value to obtain a simulation calculation result;

[0118] If the simulation calculation result meets the preset risk control target, the initial adjustment condition set is determined as the adjustment condition set;

[0119] If the simulation calculation result does not meet the preset risk control target, the condition parameters of the condition set are iteratively optimized to obtain an adjusted condition set.

[0120] In an embodiment of the present invention, the extraction refers to the process of screening out historical claims data from multidimensional risk data, the comparative analysis refers to comparing and analyzing the extracted historical claims data with the initial control values to reveal the relationship, differences and potential laws between the two, and provide valuable information for risk management, the simulation calculation refers to the process of performing numerical calculation and simulation analysis based on a given initial adjustment condition set and initial control value by establishing a model or applying a specific algorithm, and the iterative optimization of condition parameters refers to the process of iteratively adjusting the parameters in the condition set according to the simulation calculation results.

[0121] Specifically, historical claims data is filtered out from multidimensional risk data based on specific identifiers, fields, or conditions. For example, through database query statements, historical claims data that meets the conditions is extracted based on relevant fields such as the time range, claim type, and policy number of the claim record to form an independent data set. Various statistical indicators of the historical claims data, such as the mean, median, standard deviation, maximum value, and minimum value, are calculated and numerically compared with the initial control value. At the same time, the distribution and change trends of historical claims data and initial control values are intuitively displayed by drawing charts (such as bar charts, line charts, box plots, etc.). Correlation analysis and other methods can also be used to calculate the correlation coefficient between the two to quantify the degree of linear relationship between them. Finally, based on these comparisons and analyses, the comparative analysis results are summarized, including the differences between the two, whether the trends are consistent, and the strength of the correlation.

[0122] Furthermore, the comparative analysis results are carefully studied to determine what implications the relationship between historical claims data and the initial control values has for the parameters of the conditional set. For example, if the mean of the historical claims data is found to be much higher than the initial control value, this may mean that the weight setting of the key risk data is unreasonable, or the threshold is too high, and corresponding adjustments are needed. Based on the interpretation of the comparative analysis results, preliminary adjustments are made to the parameters in the conditional set. The weights of the key risk data may be adjusted, increasing the weight of risk data with a high correlation with historical claims data, and reducing the weight of risk data with a low correlation. The coefficients describing the relationship between key risk data and risk may also be adjusted to make them more consistent with the actual situation reflected by the historical claims data. In addition, based on the distribution characteristics of the historical claims data, the threshold may be appropriately raised or lowered to more accurately divide the risk level.

[0123] Furthermore, the parameters and initial control values in the initial adjustment condition set are substituted into a pre-set simulation calculation model or algorithm. The model or algorithm will perform numerical calculations based on the input parameters, simulate the risk assessment under given conditions, obtain simulation calculation results, and compare the simulation calculation results with the preset risk control objectives. The preset risk control objectives can be a specific risk value, a risk level range, a fit index with historical data, or other standards related to risk management objectives. If the simulation calculation results meet the expected goals, it means that the initial adjustment condition set is valid and can enable the risk assessment results to meet the expected requirements. At this time, the initial adjustment condition set can be determined as the adjustment condition set for actual risk management decisions.

[0124] When simulation results fail to meet the pre-set risk control objectives, in-depth analysis of the causes of the deviation is necessary. This could be due to inaccurate parameter adjustments in the initial adjustment condition set, which fail to fully account for the complex relationships between historical claims data and other risk factors. Alternatively, the simulation model itself may have limitations and fail to accurately reflect the actual risk situation. Based on this analysis of the causes of the deviation, further optimization and adjustment of the parameters in the condition set are performed. This is the iterative optimization process. More refined adjustment strategies can be employed, such as gradually fine-tuning weights, coefficients, or thresholds, observing the changing trends in the simulation results to find a parameter combination that brings the results closer to the expected target. Other relevant factors or data can also be introduced to expand and refine the condition set to improve the model's accuracy and adaptability. After iterative parameter optimization, the adjusted condition set is simulated again with the initial control values, and the new simulation results are compared with the expected target. This process is repeated until the simulation results meet the pre-set risk control objectives. The resulting condition set is then the final adjusted condition set.

[0125] In this embodiment of the present invention, extracting historical claims data from multidimensional risk data allows for a focus on actual claims settlements. This data directly reflects the actual outcomes of past risk events. Comparing this data with initial control values clarifies the discrepancy between the currently set control values and the historically actual risk profile, helping to accurately understand the actual level and changing trends of risk, and avoiding the disconnect between setting control values based solely on theory or experience and the actual situation.

[0126] In the embodiment of the present invention, by optimizing the condition set based on the initial control value, the risk assessment model can be adjusted in a timely manner according to new risk data and business conditions, so that the risk management system maintains dynamic adaptability.

[0127] S6. Calculate the target control value of the multi-dimensional risk data according to the adjustment condition set and the key risk data.

[0128] Specifically, the adjustment condition set is obtained after continuous adjustment and optimization of the initial condition set. It can more accurately reflect the risk status and the relationship between various factors. The key risk data is the core data that has an important impact on risk assessment selected from the multidimensional risk data. By combining the various parameters and rules in the adjustment condition set with the key risk data, and using the corresponding mathematical model or algorithm for calculation, we can finally obtain a target control value that can comprehensively reflect the overall risk level of the multidimensional risk data and can be used as a control standard, that is, the final premium.

[0129] It can be seen that in the above scheme, for the target control value business, the insured's multidimensional risk data is obtained, and the multidimensional risk data is cross-extracted with implicit correlation features to obtain feature correlation data; the comprehensive risk index of the multidimensional risk data is determined based on the feature correlation data, and the comprehensive risk index is analyzed with the multidimensional risk data for correlation to obtain analysis results; data extraction is performed on the multidimensional risk data based on the analysis results to obtain key risk data; a condition set is constructed based on the key risk data, and the initial control value of the multidimensional risk data is calculated based on the condition set and the key risk data; the condition parameters of the condition set are optimized based on the initial control value to obtain an adjusted condition set; the target control value of the multidimensional risk data is calculated based on the adjusted condition set and the key risk data, and through the cross-extraction of implicit correlation features, a richer basis is provided for risk assessment and decision-making, and at the same time, the condition set is adjusted and optimized to keep the risk management system dynamically adaptable.

[0130] 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 invention.

[0131] In one embodiment, a device for analyzing the insurance control value of an insured is provided. The device for analyzing the insurance control value of an insured corresponds to the method for analyzing the insurance control value of an insured in the above embodiment. Figure 3 As shown, the device for analyzing the policyholder's insurance control value includes an acquisition module 101, an extraction module 102, an indicator determination module 103, an analysis module 104, a data extraction module 105, a construction module 106, an initial calculation module 107, an optimization module 108, and a target calculation module 109. Detailed descriptions of each functional module are as follows:

[0132] Acquisition module 101, for acquiring multi-dimensional risk data of the insured;

[0133] Extraction module 102, configured to perform cross-extraction of implicit correlation features on the multi-dimensional risk data to obtain feature correlation data;

[0134] An indicator determination module 103 is configured to determine a comprehensive risk indicator of the multi-dimensional risk data based on the feature association data;

[0135] An analysis module 104 is configured to perform a correlation analysis between the comprehensive risk indicator and the multi-dimensional risk data to obtain an analysis result;

[0136] An extraction module 105 is configured to extract the multi-dimensional risk data based on the analysis results to obtain key risk data;

[0137] A construction module 106 is configured to construct a condition set based on the key risk data;

[0138] An initial calculation module 107, configured to calculate an initial control value of the multi-dimensional risk data according to the condition set and the key risk data;

[0139] The optimization module 108 is configured to optimize the condition parameters of the condition set according to the initial control value to obtain an adjusted condition set.

[0140] The target calculation module 109 calculates the target control value of the multi-dimensional risk data according to the adjustment condition set and the key risk data.

[0141] In one embodiment, when the extraction module 102 performs cross-extraction of implicit correlation features on the multi-dimensional risk data to obtain feature correlation data, it is configured to:

[0142] Performing a rough risk feature screening on the multi-dimensional risk data to obtain retained feature data;

[0143] Calculating a correlation coefficient of the retained feature data, and performing redundant feature recognition on the retained feature data according to the correlation coefficient to obtain recognition feature data;

[0144] Performing high-order cross-combination on the identification feature data to obtain combined data;

[0145] Performing correlation feature extraction on the combined data to obtain feature correlation data.

[0146] In one embodiment, when determining the comprehensive risk index of the multi-dimensional risk data based on the feature association data, the index determination module 103 is configured to:

[0147] Evaluate the importance of each feature-associated data in multidimensional risk data;

[0148] According to the importance of the feature-related data, weight configuration is performed on the feature-related data one by one to obtain the weight corresponding to the feature-related data;

[0149] The weight is weighted and summed with the feature-related data to obtain a comprehensive risk index.

[0150] In one embodiment, when the analysis module 104 performs a correlation analysis on the comprehensive risk indicator and the multi-dimensional risk data to obtain an analysis result, it is configured to:

[0151] Calculating the Pearson correlation coefficient between the comprehensive risk indicator and each dimension data in the multidimensional risk data to obtain a coefficient result;

[0152] A correlation analysis is performed on the comprehensive risk indicator and the multidimensional risk data according to the coefficient result to obtain an analysis result.

[0153] In one embodiment, when constructing a condition set based on the key risk data, the construction module 106 is configured to:

[0154] Objectively weighting the key risk data according to the statistical characteristics of the data to obtain the weight corresponding to the key risk data;

[0155] Performing linear regression analysis on the key risk data to obtain coefficients corresponding to the key risk data;

[0156] Performing statistical distribution analysis on the key risk data, and determining thresholds corresponding to the key risk data based on the statistical distribution analysis results;

[0157] A condition set is generated based on the weights, coefficients and thresholds.

[0158] In one embodiment, when calculating the initial control value of the multi-dimensional risk data according to the condition set and the key risk data, the initial calculation module 107 is configured to:

[0159] Calculating a risk assessment value of the key risk data according to the condition set;

[0160] determining a fuzzy initial control value according to the risk assessment value;

[0161] Linearly combining the weights and coefficients in the condition set to obtain a combination coefficient;

[0162] An initial control value of the multidimensional risk data is calculated according to the fuzzy initial control value and the combination coefficient.

[0163] In one embodiment, when optimizing the condition parameters of the condition set according to the initial control value to obtain the adjusted condition set, the optimization module 108 is configured to:

[0164] Extracting historical claims data from the multidimensional risk data, and performing comparative analysis on the historical claims data and the initial control value to obtain comparative analysis results;

[0165] Performing initial adjustment on condition parameters of the condition set according to the comparative analysis result to obtain an initial adjusted condition set;

[0166] Performing simulation calculation on the initial adjustment condition set and the initial control value to obtain a simulation calculation result;

[0167] If the simulation calculation result meets the preset risk control target, the initial adjustment condition set is determined as the adjustment condition set;

[0168] If the simulation calculation result does not meet the preset risk control target, the condition parameters of the condition set are iteratively optimized to obtain an adjusted condition set.

[0169] The present invention provides an insurance control value analysis device for an insured person. For a target control value business, the device obtains multidimensional risk data of the insured person, performs implicit correlation feature cross-extraction on the multidimensional risk data to obtain feature correlation data; determines a comprehensive risk index of the multidimensional risk data based on the feature correlation data, and performs correlation analysis on the comprehensive risk index and the multidimensional risk data to obtain an analysis result; extracts data from the multidimensional risk data based on the analysis result to obtain key risk data; constructs a condition set based on the key risk data, and calculates an initial control value of the multidimensional risk data based on the condition set and the key risk data; optimizes condition parameters of the condition set based on the initial control value to obtain an adjusted condition set; calculates a target control value of the multidimensional risk data based on the adjusted condition set and the key risk data. Through the implicit correlation feature cross-extraction, a richer basis is provided for risk assessment and decision-making. At the same time, the condition set is adjusted and optimized to ensure that the risk management system maintains dynamic adaptability.

[0170] The specific definition of a device for analyzing a policyholder's insurance control value can be found in the definition of a method for analyzing a policyholder's insurance control value above and will not be further elaborated here. Each module in the aforementioned device for analyzing a policyholder's insurance control value can be implemented in whole or in part via software, hardware, or a combination thereof. Each of these modules can be embedded in or independent of a processor in a computer device in hardware form, or stored in a computer device memory in software form, so that the processor can call and execute the corresponding operations of each module.

[0171] 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 4 As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client via a network connection. When executed by the processor, the computer program implements the functions or steps on the service side of a method for analyzing insurance control values of an insured.

[0172] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, memory, a network interface, a display screen, and an input device connected via a system bus. 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 internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When executed by the processor, the computer program implements the client-side functions or steps of a method for analyzing a policyholder's insurance control value.

[0173] 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. When the processor executes the computer program, the following steps are performed:

[0174] Acquire multidimensional risk data of the insured, and perform cross-extraction of implicit correlation features on the multidimensional risk data to obtain feature correlation data;

[0175] Determining a comprehensive risk index of the multidimensional risk data based on the characteristic association data, and performing a correlation analysis between the comprehensive risk index and the multidimensional risk data to obtain an analysis result;

[0176] Extracting the multidimensional risk data according to the analysis results to obtain key risk data;

[0177] Constructing a condition set according to the key risk data, and calculating an initial control value of the multi-dimensional risk data according to the condition set and the key risk data;

[0178] Optimizing condition parameters of the condition set according to the initial control value to obtain an adjusted condition set;

[0179] A target control value of the multi-dimensional risk data is calculated according to the adjustment condition set and the key risk data.

[0180] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0181] Acquire multidimensional risk data of the insured, and perform cross-extraction of implicit correlation features on the multidimensional risk data to obtain feature correlation data;

[0182] Determining a comprehensive risk index of the multidimensional risk data based on the characteristic association data, and performing a correlation analysis between the comprehensive risk index and the multidimensional risk data to obtain an analysis result;

[0183] Extracting the multidimensional risk data according to the analysis results to obtain key risk data;

[0184] Constructing a condition set according to the key risk data, and calculating an initial control value of the multi-dimensional risk data according to the condition set and the key risk data;

[0185] Optimizing condition parameters of the condition set according to the initial control value to obtain an adjusted condition set;

[0186] A target control value of the multi-dimensional risk data is calculated according to the adjustment condition set and the key risk data.

[0187] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can be found in the relevant descriptions of the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0188] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. 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 can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various 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).

[0189] 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 device can be divided into different functional units or modules to complete all or part of the functions described above.

[0190] The above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. If software tools or components other than those of the company appear in the application embodiments, they are merely used for illustration and do not represent actual use. Although the present invention has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above-mentioned embodiments, or make equivalent replacements for some of the technical features therein. However, these 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 invention, and should all be included in the scope of protection of the present invention.

Claims

1. A method for analyzing the insurance control value of an insured, characterized in that: include: Acquire multidimensional risk data of the insured, and perform cross-extraction of implicit correlation features on the multidimensional risk data to obtain feature correlation data; Determining a comprehensive risk index of the multidimensional risk data based on the characteristic association data, and performing a correlation analysis between the comprehensive risk index and the multidimensional risk data to obtain an analysis result; Extracting the multidimensional risk data according to the analysis results to obtain key risk data; Constructing a condition set according to the key risk data, and calculating an initial control value of the multi-dimensional risk data according to the condition set and the key risk data; Optimizing condition parameters of the condition set according to the initial control value to obtain an adjusted condition set; A target control value of the multi-dimensional risk data is calculated according to the adjustment condition set and the key risk data.

2. The method for analyzing the policyholder's insurance control value according to claim 1, wherein: The step of performing cross-extraction of implicit correlation features on the multi-dimensional risk data to obtain feature correlation data includes: Performing a rough risk feature screening on the multi-dimensional risk data to obtain retained feature data; Calculating a correlation coefficient of the retained feature data, and performing redundant feature recognition on the retained feature data according to the correlation coefficient to obtain recognition feature data; Performing high-order cross-combination on the identification feature data to obtain combined data; Performing correlation feature extraction on the combined data to obtain feature correlation data.

3. The method for analyzing the policyholder's insurance control value according to claim 1, wherein: Determining the comprehensive risk index of the multi-dimensional risk data based on the feature association data includes: Evaluate the importance of each feature-associated data in multidimensional risk data; According to the importance of the feature-related data, weight configuration is performed on the feature-related data one by one to obtain the weight corresponding to the feature-related data; The weight is weighted and summed with the feature-related data to obtain a comprehensive risk index.

4. The method for analyzing the policyholder's insurance control value according to claim 1, wherein: The performing of correlation analysis on the comprehensive risk indicator and the multi-dimensional risk data to obtain analysis results includes: Calculating the Pearson correlation coefficient between the comprehensive risk indicator and each dimension data in the multidimensional risk data to obtain a coefficient result; A correlation analysis is performed on the comprehensive risk indicator and the multidimensional risk data according to the coefficient result to obtain an analysis result.

5. The method for analyzing the policyholder's insurance control value according to claim 4, wherein: The step of constructing a condition set according to the key risk data includes: Objectively weighting the key risk data according to the statistical characteristics of the data to obtain the weight corresponding to the key risk data; Performing linear regression analysis on the key risk data to obtain coefficients corresponding to the key risk data; Performing statistical distribution analysis on the key risk data, and determining thresholds corresponding to the key risk data based on the statistical distribution analysis results; A condition set is generated based on the weights, coefficients and thresholds.

6. The method for analyzing the policyholder's insurance control value according to claim 1, wherein: Calculating the initial control value of the multi-dimensional risk data according to the condition set and the key risk data includes: Calculating a risk assessment value of the key risk data according to the condition set; determining a fuzzy initial control value according to the risk assessment value; Linearly combining the weights and coefficients in the condition set to obtain a combination coefficient; An initial control value of the multidimensional risk data is calculated according to the fuzzy initial control value and the combination coefficient.

7. The method for analyzing the policyholder's insurance control value according to claim 1, wherein: Optimizing the condition parameters of the condition set according to the initial control value to obtain the adjusted condition set includes: Extracting historical claims data from the multidimensional risk data, and performing comparative analysis on the historical claims data and the initial control value to obtain comparative analysis results; Performing initial adjustment on condition parameters of the condition set according to the comparative analysis result to obtain an initial adjusted condition set; Performing simulation calculation on the initial adjustment condition set and the initial control value to obtain a simulation calculation result; If the simulation calculation result meets the preset risk control target, the initial adjustment condition set is determined as the adjustment condition set; If the simulation calculation result does not meet the preset risk control target, the condition parameters of the condition set are iteratively optimized to obtain an adjusted condition set.

8. A device for analyzing the insurance control value of an insured, characterized in that: include: Acquisition module, used to obtain the policyholder's multi-dimensional risk data; An extraction module, configured to perform cross-extraction of implicit correlation features on the multi-dimensional risk data to obtain feature correlation data; An indicator determination module, configured to determine a comprehensive risk indicator of the multi-dimensional risk data based on the feature association data; An analysis module, configured to perform a correlation analysis on the comprehensive risk indicator and the multi-dimensional risk data to obtain an analysis result; A data extraction module, configured to extract data from the multidimensional risk data according to the analysis results to obtain key risk data; A construction module, configured to construct a condition set based on the key risk data; an initial calculation module, configured to calculate an initial control value of the multidimensional risk data based on the condition set and the key risk data; an optimization module, configured to optimize condition parameters of the condition set according to the initial control value to obtain an adjusted condition set; A target calculation module calculates a target control value of the multi-dimensional risk data according to the adjustment condition set and the key risk data.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for analyzing the insurance control value of the insured according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for analyzing the insurance control value of the insured according to any one of claims 1 to 7 is implemented.

Images