Self-adaptive scoring method and system suitable for multiple injury condition assessment criteria

Abstract

The invention provides a self-adaptive scoring method and system suitable for various injury assessment standards. The method comprises the following steps: receiving injury condition evaluation standards selected by a user, performing comprehensive consideration, and generating a customized scoring rule set; based on the customized scoring rule set, applying a symbolic regression algorithm, optimizing scoring model performance, capturing a nonlinear relationship in data, adopting a random matrix theory technology, analyzing and identifying an abnormal value and a potential mode, and generating a scoring adjustment coefficient; on the basis of the score adjustment coefficient, a quantum enhancement learning algorithm is applied, the powerful parallel processing capacity of quantum computing is utilized to accelerate model training, a causal reasoning technology is adopted, a causal graph is constructed to execute an intervention simulation experiment, and a correction score report is generated; based on the corrected score report, a highly interactive user interface is provided, and an adaptive score adjustment scheme is generated. The technical scheme provided by the invention is suitable for various injury assessment standards, and the flexibility and accuracy of the adaptive scoring method are improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of adaptive technology, and in particular to an adaptive scoring method and system applicable to multiple injury assessment standards. Background Art

[0002] With the prevalence of complex assessment systems, there is a growing demand for adaptive scoring methods, especially when dealing with diverse and dynamically changing scoring criteria. Different application scenarios (such as insurance claims, risk assessment, performance appraisal, etc.) require the system to be able to receive the injury assessment criteria selected by the user, and comprehensively consider the scoring dimensions and weights based on these criteria to generate a customized set of scoring rules. In addition, the system also needs to have efficient data processing capabilities, be able to analyze a large amount of relevant case data in a short period of time, and ensure the accuracy and robustness of the scoring results.

[0003] Currently, most scoring systems use traditional machine learning algorithms and static rule sets for evaluation. These systems usually process data through predefined scoring models and fixed parameter settings, such as linear regression or decision tree models. In the initial scoring stage, the system is trained based on historical data and scores new cases using fixed rules. For outlier identification and pattern mining, existing solutions mostly rely on traditional statistical methods, such as principal component analysis and cluster analysis.

[0004] However, existing scoring methods have significant limitations. Traditional scoring systems have difficulty coping with changes in multiple evaluation criteria, cannot dynamically adjust scoring rules according to different application scenarios, and are not flexible enough for adaptive scoring methods. In addition, existing methods perform poorly in dealing with nonlinear relationships and complex feature interactions, resulting in inaccurate scoring results. At the same time, traditional algorithms have low computational efficiency when faced with large-scale data and cannot fully utilize parallel computing resources to accelerate model training. More importantly, existing systems lack in-depth causal reasoning capabilities and have difficulty revealing the independent impact of key factors, which limits the transparency and interpretability of the scoring system. Therefore, there is an urgent need for a more flexible, efficient, and accurate adaptive scoring method to meet diverse evaluation needs. Summary of the invention

[0005] The embodiments of the present application provide an adaptive scoring method and system applicable to a variety of injury assessment standards, so as to solve the problem of insufficient flexibility of the adaptive scoring method in the prior art.

[0006] In a first aspect, the present application provides an adaptive scoring method applicable to multiple injury assessment criteria, including:

[0007] Receive the injury assessment criteria selected by the user, comprehensively consider the scoring dimensions and weights under different assessment criteria, and generate a customized scoring rule set;

[0008] Based on the customized scoring rule set, a symbolic regression algorithm is used to retrieve relevant injury case data for preliminary scoring, and the scoring model performance is optimized by searching for the optimal injury change expression, capturing the nonlinear relationship in the data, and using random matrix theory technology to analyze and identify outliers and potential patterns, thereby improving the accuracy and robustness of the scoring results and generating a scoring adjustment coefficient;

[0009] Based on the score adjustment coefficient, the quantum enhanced learning algorithm is used to accelerate model training by taking advantage of the powerful parallel processing capabilities of quantum computing, explore a large number of possible solution spaces, use causal reasoning technology for in-depth analysis, build a causal graph to perform intervention simulation experiments, reveal the independent impact of key factors, and generate a correction score report;

[0010] Based on the correction score report, a highly interactive user interface is provided, which combines user feedback to allow the user to make flexible adjustments according to actual conditions and generate an adaptive score adjustment plan.

[0011] Optionally, based on the customized scoring rule set, a symbolic regression algorithm is used to retrieve relevant injury case data for preliminary scoring, the scoring model performance is optimized by searching for the optimal injury change expression, the nonlinear relationship in the data is captured, and the random matrix theory technology is used to analyze and identify outliers and potential patterns, improve the accuracy and robustness of the scoring results, and generate a scoring adjustment coefficient, including:

[0012] Based on the customized scoring rule set, the injury case data in the database is screened and retrieved to match the injury assessment criteria selected by the user, and a relevant injury case data set is generated;

[0013] Based on the relevant injury case data set, a symbolic regression algorithm is used to perform preliminary scoring calculations on each case, and by searching for the optimal injury change expression, modeling the correlation between different injury characteristics, and automatically exploring possible function combinations, a basic scoring result is generated;

[0014] Based on the basic scoring results, random matrix theory techniques are used to statistically analyze the distribution of eigenvalues ​​of the data covariance matrix, identify outliers and potential patterns, and generate an outlier correction matrix;

[0015] Based on the outlier correction matrix, the abnormal situation is accurately adjusted to improve the accuracy and robustness of the scoring results, and the scoring adjustment coefficient is generated.

[0016] Optionally, based on the relevant injury case data set, a symbolic regression algorithm is used to perform preliminary scoring calculations on each case, and by searching for the optimal injury change expression, modeling the correlation between different injury characteristics, and automatically exploring possible function combinations, a basic scoring result is generated, including:

[0017] Based on the relevant injury case data set, data cleaning and normalization processing are performed, and features related to injury assessment are identified through feature extraction to generate a pre-processed injury case data set;

[0018] Based on the preprocessed injury case data set, a symbolic regression algorithm is used to perform a preliminary score calculation on each case, and the relationship between different injury characteristics is modeled by searching for the optimal injury change expression to generate a preliminary scoring model for each case;

[0019] Based on the preliminary scoring model for each case, further analyze the correlation between different injury characteristics, identify significant associations to quantify the degree of association, and generate a feature correlation analysis report;

[0020] Based on the feature correlation analysis report, the scoring model performance is optimized to ensure that the interactions between multiple features are comprehensively considered to generate basic scoring results.

[0021] Optionally, based on the basic scoring results, random matrix theory techniques are used to perform statistical analysis on the distribution of eigenvalues ​​of the data covariance matrix, identify outliers and potential patterns, and generate an outlier correction matrix, including:

[0022] Based on the basic scoring results, combined with key features in the original injury case data, new features are introduced or existing feature weights are adjusted to generate a feature enhanced data set;

[0023] Based on the feature enhancement data set, the covariance matrix is ​​calculated using random matrix theory technology, and the eigenvalue distribution is statistically analyzed. The nonlinear relationship in the data is captured through random matrix theory to generate an optimized basic scoring result;

[0024] Based on the optimized basic scoring results, further analyze the correlation between different injury characteristics, identify the characteristic values ​​that significantly deviate from the normal distribution and the potential patterns hidden behind the data, and generate an outlier and potential pattern recognition report;

[0025] Based on the outlier and potential pattern recognition report, weights and adjustment factors are introduced to correct the scoring deviation caused by outliers and generate an outlier correction matrix.

[0026] Optionally, based on the score adjustment coefficient, a quantum enhanced learning algorithm is used to accelerate model training by using the powerful parallel processing capability of quantum computing, explore a large number of possible solution spaces, use causal reasoning technology for in-depth analysis, build a causal graph to perform intervention simulation experiments, reveal the independent impact of key factors, and generate a correction score report, including:

[0027] Based on the score adjustment coefficient, make fine adjustments, recalibrate various key parameters, and generate an optimized score model;

[0028] Based on the optimized scoring model, the quantum enhanced learning algorithm is used to accelerate model training by taking advantage of the powerful parallel processing capabilities of quantum computing. Through the superposition and entanglement characteristics of quantum bits, a large number of possible solution spaces can be explored in a short period of time to generate multiple possible scoring schemes.

[0029] Based on the multi-possibility scoring scheme, causal reasoning technology is used to perform correlation analysis, explore the causal relationship between variables, and construct a causal diagram to perform intervention simulation experiments, reveal the independent impact of key factors, and generate a detailed causal relationship analysis report;

[0030] Based on the detailed causal analysis report, the specific impact of each key factor is explained in detail to ensure the flexibility and transparency of the scoring system and generate a corrected scoring report.

[0031] Optionally, the scoring model based on the optimization uses a quantum enhanced learning algorithm, utilizes the powerful parallel processing capability of quantum computing to accelerate model training, and explores a large number of possible solution spaces in a short time through the superposition and entanglement state characteristics of quantum bits to generate multiple possible scoring schemes, including:

[0032] Based on the optimized scoring model, cross-validation and small-scale test set application are used to ensure stability and consistency on different data sets, and generate a consistent scoring model;

[0033] Based on the consistency scoring model, the quantum enhanced learning algorithm is used to accelerate model training by taking advantage of the powerful parallel processing capabilities of quantum computing. Through the superposition and entanglement characteristics of quantum bits, a large number of possible solution spaces can be explored in a short period of time to generate a scoring model for accelerated training.

[0034] Based on the accelerated training scoring model, a high-dimensional feature space is constructed, and a variety of injury characteristics and their interrelationships are comprehensively considered to generate a high-dimensional feature representation;

[0035] Based on the high-dimensional feature representation, multiple potential changes are comprehensively considered and multiple possible scoring schemes are generated by simulating different condition scenarios.

[0036] Optionally, based on the correction score report, a highly interactive user interface is provided, which allows the user to make flexible adjustments according to actual conditions in combination with user feedback, and generates an adaptive score adjustment scheme, including:

[0037] Based on the correction scoring report, the correction scoring results are graphically displayed, detailed scoring basis and key factor impact analysis are provided, and a highly interactive user interface is generated;

[0038] Based on the highly interactive user interface, it allows users to make flexible adjustments according to the actual situation, collect user feedback information in real time, and generate a user feedback dataset;

[0039] Based on the user feedback dataset, the system combines historical data with the current user input, and uses the Bayesian update method to improve flexibility and adaptability, and generates an adaptive scoring model;

[0040] Based on the adaptive scoring model, deeply integrate all feedback information during the interaction process, ensure the transparency of the scoring system and user participation, and generate an adaptive scoring adjustment plan.

[0041] In a second aspect, an embodiment of the present application provides an adaptive scoring system applicable to multiple injury assessment criteria, including:

[0042] A receiving module, configured to receive the injury assessment criteria selected by the user, comprehensively consider the scoring dimensions and weights under different assessment criteria, and generate a customized scoring rule set;

[0043] A retrieval module, configured to, based on the customized scoring rule set, use the symbolic regression algorithm to retrieve relevant injury case data for preliminary scoring, optimize the performance of the scoring model by searching for the optimal injury change expression, capture the non-linear relationship in the data, adopt the random matrix theory technology, analyze and identify outliers and potential patterns, improve the accuracy and robustness of the scoring result, and generate a scoring adjustment coefficient;

[0044] An analysis module, configured to, based on the scoring adjustment coefficient, use the quantum enhanced learning algorithm, utilize the powerful parallel processing ability of quantum computing to accelerate model training, explore a large number of possible solution spaces, adopt causal reasoning technology for in-depth analysis, construct a causal graph to perform intervention simulation experiments, reveal the independent influence of key factors, and generate a corrected scoring report;

[0045] A generation module, configured to, based on the corrected scoring report, provide a highly interactive user interface, combine user feedback, allow users to make flexible adjustments according to the actual situation, and generate an adaptive scoring adjustment plan.

[0046] In a third aspect, an embodiment of the present application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement an adaptive scoring method applicable to multiple injury assessment criteria as described in the first aspect.

[0047] In a fourth aspect, an embodiment of the present application provides a computer storage medium, storing a computer program, and when the computer program is executed by a computer, it implements an adaptive scoring method applicable to multiple injury assessment criteria as described in the first aspect.

[0048] In an embodiment of the present application, an injury assessment standard selected by a user is received, and the scoring dimensions and weights under different assessment standards are comprehensively considered to generate a customized scoring rule set; based on the customized scoring rule set, a symbolic regression algorithm is used to retrieve relevant injury case data for preliminary scoring, and the scoring model performance is optimized by searching for the optimal injury change expression, capturing nonlinear relationships in the data, and using random matrix theory technology to analyze and identify outliers and potential patterns, improve the accuracy and robustness of the scoring results, and generate a scoring adjustment coefficient; based on the scoring adjustment coefficient, a quantum enhanced learning algorithm is used, and the powerful parallel processing capabilities of quantum computing are used to accelerate model training, explore a large number of possible solution spaces, use causal reasoning technology for in-depth analysis, construct a causal diagram to perform intervention simulation experiments, reveal the independent influence of key factors, and generate a corrected scoring report; based on the corrected scoring report, a highly interactive user interface is provided, combined with user feedback, allowing the user to make flexible adjustments based on actual conditions, and generate an adaptive scoring adjustment plan. It receives the injury assessment criteria selected by the user, comprehensively considers the scoring dimensions and weights according to different assessment criteria, and generates a customized set of scoring rules. This method ensures the flexibility and adaptability of the scoring system to meet the needs of different application scenarios. It optimizes the performance of the scoring model through symbolic regression algorithm, captures nonlinear relationships in the data, and uses random matrix theory technology to analyze and identify outliers and potential patterns, which significantly improves the accuracy and robustness of the scoring results. It uses the powerful parallel processing capability of the quantum reinforcement learning algorithm to explore a large number of possible solution spaces, significantly shortens the model training time, and improves the degree of model optimization. It provides a highly interactive user interface, combined with user feedback, allowing users to flexibly adjust the scoring scheme according to actual conditions, thereby enhancing user participation and satisfaction.

[0049] Furthermore, by screening and retrieving injury case data in the database, we ensure that the case data used is highly matched with the evaluation criteria selected by the user, generate relevant injury case data sets, and provide a high-quality data basis for subsequent scoring calculations; use symbolic regression algorithms to automatically explore the optimal injury change expression, model the correlation between different injury characteristics, reduce manual intervention, and improve the automation and efficiency of scoring calculations; use random matrix theory technology to perform statistical analysis on the eigenvalue distribution of the data covariance matrix, effectively identify outliers and potential patterns, generate an outlier correction matrix, and ensure the reliability of the scoring results; accurately adjust abnormal situations based on the outlier correction matrix, further improve the accuracy and robustness of the scoring results, and generate reliable scoring adjustment coefficients.

[0050] Furthermore, based on the scoring adjustment coefficient, fine-tuning is performed, key parameters are recalibrated, and an optimized scoring model is generated to ensure that the scoring model can more accurately reflect the actual situation; the powerful parallel processing capability of the quantum reinforcement learning algorithm is used to accelerate model training, explore a large number of possible solution spaces in a short period of time, and generate multiple possible scoring schemes, thereby improving the model training efficiency and exploration scope; causal reasoning technology is used to perform correlation analysis, explore the causal relationship between variables, and perform intervention simulation experiments by constructing causal graphs to reveal the independent influence of key factors, generate detailed causal analysis reports, and enhance the transparency and interpretability of the scoring system; based on the detailed causal analysis report, the specific impact of each key factor is explained in detail to ensure that the scoring system is highly flexible and transparent, and finally a correction scoring report is generated to provide users with a clear scoring basis and reference.

[0051] These and other aspects of the present application will be more clearly understood in the description of the following embodiments; BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0053] Figure 1 A flowchart of an adaptive scoring method applicable to multiple injury assessment criteria provided in an embodiment of the present application;

[0054] Figure 2 A schematic diagram of the structure of an adaptive scoring system applicable to multiple injury assessment standards provided in an embodiment of the present application;

[0055] Figure 3 A schematic diagram of the structure of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0056] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0057] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0058] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0059] Figure 1 A flowchart of an adaptive scoring method applicable to multiple injury assessment criteria is provided for the embodiment of the present application, such as Figure 1 As shown, the method includes:

[0060] 101. Receive the injury assessment criteria selected by the user, comprehensively consider the scoring dimensions and weights under different assessment criteria, and generate a customized scoring rule set;

[0061] In this step, the injury assessment criteria include various types of data, such as medical diagnosis codes, treatment options, recovery time, etc., which are used to define the specific requirements and goals of the scoring.

[0062] Scoring dimensions refer to the various factors considered in the scoring process, such as injury severity, injury location, treatment method, etc.

[0063] A customized scoring rule set is a scoring guide that is dynamically generated based on the injury assessment criteria selected by the user, ensuring that the scoring system can adapt to different application scenarios.

[0064] In an embodiment of the present application, the injury assessment criteria selected by the user are first received, and then the scoring dimensions and weights under different assessment criteria are comprehensively considered, and the weights of each dimension are calculated using a built-in algorithm, and finally a customized scoring rule set is generated to provide a basis for subsequent scoring calculations.

[0065] Suppose the user selects bodily injury caused by a traffic accident as the evaluation criterion in the insurance claim system; first, the system automatically identifies and loads relevant scoring dimensions based on the user's selected criteria, such as fracture type, injury site, number of hospitalization days and other data, for preliminary scoring; secondly, the system calculates the weight of each dimension through a built-in algorithm to ensure the accuracy and fairness of the scoring results; thirdly, the system generates a customized set of scoring rules to provide a basis for subsequent scoring calculations.

[0066] 102. Based on the customized scoring rule set, use the symbolic regression algorithm to retrieve relevant injury case data for preliminary scoring, optimize the scoring model performance by searching for the optimal injury change expression, capture the nonlinear relationship in the data, use random matrix theory technology to analyze and identify outliers and potential patterns, improve the accuracy and robustness of the scoring results, and generate a scoring adjustment coefficient;

[0067] In this step, the symbolic regression algorithm is an evolutionary algorithm that aims to optimize the scoring model performance by searching for the best mathematical expression to describe data features. The algorithm finds the expression that most accurately reflects the relationship between input data and output results through continuous iteration and improvement.

[0068] The optimal injury change expression refers to the best mathematical formula found by the symbolic regression algorithm to describe the complex relationship between different injury characteristics. It can capture the nonlinear relationship in the data and provide more accurate results in the scoring process.

[0069] A nonlinear relationship means that the relationship between two or more variables is not a simple linear proportional relationship, but a complex, indirect proportional relationship. These relationships may include exponential, logarithmic, polynomial and other forms. The symbolic regression algorithm can effectively capture this complex relationship.

[0070] Random matrix theory technology is a statistical method used to analyze and identify outliers and potential patterns. Through statistical analysis of the distribution of eigenvalues ​​of the data covariance matrix, outliers and potential patterns in the data can be identified, thereby improving the accuracy and robustness of the scoring results.

[0071] Scoring adjustment coefficients are a set of adjustment factors generated based on data analysis results, which are used to correct preliminary scoring results and improve the accuracy of the final scoring. These coefficients reflect the sensitivity and impact of the scoring model on specific data features.

[0072] In the embodiment of the present application, the system uses a symbolic regression algorithm based on a customized scoring rule set to retrieve and preliminarily score relevant injury case data in the database, optimizes the scoring model performance by searching for the optimal injury change expression, and uses random matrix theory technology to analyze and identify outliers and potential patterns to generate a scoring adjustment coefficient.

[0073] For example, continuing with the above example, assume that the user has selected physical injuries caused by traffic accidents as the evaluation criteria; first, the system retrieves injury cases similar to the current case from the historical claims database based on a customized set of scoring rules; second, the system uses a symbolic regression algorithm to perform preliminary scoring on these cases, automatically exploring the optimal expression of injury changes and capturing nonlinear relationships in the data; third, the system uses random matrix theory technology to analyze the preliminary scoring results and identify possible outliers and potential patterns; finally, the system generates a scoring adjustment coefficient to prepare for the next step of in-depth optimization.

[0074] 103. Based on the score adjustment coefficient, the quantum enhanced learning algorithm is used to accelerate model training by taking advantage of the powerful parallel processing capability of quantum computing, explore a large number of possible solution spaces, use causal reasoning technology for in-depth analysis, construct a causal graph to perform intervention simulation experiments, reveal the independent impact of key factors, and generate a correction score report;

[0075] In this step, the quantum enhanced learning algorithm combines the powerful parallel processing capabilities of quantum computing with the advantages of traditional machine learning, and is able to explore a large number of possible solution spaces in a short period of time.

[0076] Quantum computing utilizes the superposition and entanglement properties of quantum bits to process multiple possibilities at the same time, greatly improving computing efficiency.

[0077] Parallel processing capability refers to the ability of a computer system to handle multiple tasks simultaneously.

[0078] Causal reasoning technology is a method used to reveal the causal relationship between variables. By constructing causal diagrams and conducting intervention simulation experiments, the independent influence of key factors can be revealed. This method not only improves the transparency and interpretability of the scoring system, but also ensures the rationality and credibility of the scoring results.

[0079] A causal diagram is a graphical representation method used to show the cause-effect relationship between variables. Through the causal diagram, we can clearly see which factors directly affect the scoring results, which factors have an indirect impact, and the interactions between them.

[0080] The intervention simulation experiment is based on the causal diagram. By changing the values ​​of certain variables and observing their impact on other variables, the independent influence of key factors is revealed. This method helps to understand the specific impact of different factors on the scoring results and provides transparency and interpretability of the scoring system.

[0081] The corrected score report records the specific impact of each key factor in detail, provides transparency and explanation of the scoring system, and ensures that the scoring results are more accurate and reliable. The report includes the weight of each factor, independent impact and other relevant information.

[0082] In an embodiment of the present application, the system starts a quantum enhanced learning algorithm to perform model training based on the score adjustment coefficient, uses the superposition and entanglement characteristics of quantum bits to quickly explore a large number of possible solution spaces, and uses causal reasoning technology to construct a causal graph and perform intervention simulation experiments to reveal the independent influence of key factors and generate a detailed correction score report.

[0083] For example, continuing with the above example, assume that the user has generated a score adjustment coefficient; first, the system uses the score adjustment coefficient to start the quantum enhanced learning algorithm for model training; second, the system uses the superposition and entanglement characteristics of quantum bits to quickly explore a large number of possible solution spaces; third, the system uses causal reasoning technology to construct a causal graph and perform intervention simulation experiments to reveal the independent influence of key factors; finally, the system generates a detailed correction score report, detailing the specific impact of each key factor to ensure the transparency and interpretability of the scoring system.

[0084] 104. Based on the correction score report, a highly interactive user interface is provided, which combines user feedback to allow the user to make flexible adjustments according to actual conditions and generate an adaptive score adjustment plan.

[0085] In this step, the highly interactive user interface is an intuitive and easy-to-operate platform that enables users to view detailed information in the scoring report and flexibly adjust the scoring results according to actual conditions.

[0086] The user interface includes visualization tools and operation options to help users understand the weight of each scoring dimension and the impact of key factors.

[0087] User feedback refers to the opinions and suggestions provided by users when using the rating system, including their satisfaction with the rating results, areas that need improvement, etc.

[0088] The adaptive scoring adjustment scheme is a set of scoring rules adjusted according to user feedback and actual needs, ensuring that the scoring results are closer to the actual application environment and improving user experience and satisfaction. The scheme can be continuously updated and optimized based on new data and user feedback.

[0089] The weight of the scoring dimension refers to the importance of each scoring dimension in the overall scoring. Through the user interface, users can adjust these weights according to actual conditions to make the scoring results more in line with actual application scenarios.

[0090] The impact of key factors refers to those factors that have a significant impact on the scoring results. Users can view the specific impact of these factors through the user interface and make adjustments as needed to ensure the rationality and accuracy of the scoring results.

[0091] In an embodiment of the present application, a highly interactive user interface is provided based on the correction score report, through which the user can view the detailed score report and flexibly adjust the score results according to the actual situation, dynamically adjust the score model based on user feedback, and finally generate an adaptive score adjustment plan.

[0092] For example, continuing with the above example, assume that the user has obtained a detailed correction score report; first, the system provides an intuitive and easy-to-operate user interface based on the correction score report; second, the user views the detailed score report through the interface to understand the weight of each score dimension and the impact of key factors; third, the user flexibly adjusts the score results according to actual conditions, such as modifying the weight of certain dimensions or adding new score rules; finally, the system generates an adaptive score adjustment plan based on user feedback to ensure that the score results are closer to actual needs.

[0093] In summary, steps 101 to 104 cover the complete process from the user selecting the injury assessment criteria to generating an adaptive scoring adjustment plan, aiming to provide a flexible, efficient and accurate adaptive scoring method to meet the needs of diverse and complex application scenarios.

[0094] In order to solve the accuracy and robustness problems of the scoring model when processing complex data, in some embodiments, the method for performing preliminary scoring based on a customized scoring rule set described in step 102 includes: based on the customized scoring rule set, screening and retrieving injury case data in the database to match the injury assessment criteria selected by the user, and generating a relevant injury case data set; based on the relevant injury case data set, using a symbolic regression algorithm to perform preliminary scoring calculations on each case, searching for the optimal injury change expression, modeling the correlation between different injury characteristics, automatically exploring possible function combinations, and generating basic scoring results; based on the basic scoring results, using random matrix theory technology to perform statistical analysis on the distribution of eigenvalues ​​of the data covariance matrix, identifying outliers and potential patterns, and generating an outlier correction matrix; based on the outlier correction matrix, accurately adjusting abnormal situations to improve the accuracy and robustness of the scoring results, and generating a scoring adjustment coefficient.

[0095] In this embodiment, screening and retrieval means that the system screens out injury case data similar to the current case from the database according to the injury assessment criteria selected by the user. These data are used to generate relevant injury case data sets to ensure that subsequent scoring calculations are based on high-quality data.

[0096] The relevant injury case data sets are a set of case data that are generated through screening and retrieval and are highly matched with the injury assessment criteria selected by the user. These data sets contain injury characteristics and historical scoring results that are closely related to the current assessment criteria, providing a basis for subsequent preliminary scoring calculations.

[0097] The basic scoring results are obtained after preliminary scoring calculations for each case, reflecting the basic scoring situation of each case. These scoring results are the basis for subsequent analysis and provide a preliminary scoring framework to help identify which factors have a significant impact on the score.

[0098] The outlier correction matrix is ​​a set of adjustment factors generated based on the data analysis results, which is used to correct the outliers in the preliminary scoring results. The matrix makes adjustments based on the data outliers identified through statistical analysis, which improves the accuracy and robustness of the scoring results.

[0099] In the embodiment of the present application, first, the system screens and retrieves the injury case data in the database based on a customized scoring rule set to match the injury assessment criteria selected by the user and generate a relevant injury case data set; secondly, the system uses a symbolic regression algorithm to perform a preliminary scoring calculation on each case, models the correlation between different injury characteristics by searching for the optimal injury change expression, automatically explores possible function combinations, and generates a basic scoring result; thirdly, the system uses random matrix theory technology to perform statistical analysis on the eigenvalue distribution of the data covariance matrix, identifies outliers and potential patterns, and generates an outlier correction matrix; finally, the system accurately adjusts abnormal situations based on the outlier correction matrix, improves the accuracy and robustness of the scoring results, and generates a scoring adjustment coefficient.

[0100] Here is a specific example:

[0101] Assume that the user selects a specific type of occupational disease as the evaluation criterion in the work injury claims system; first, the system screens and retrieves the injury case data in the database based on a customized scoring rule set to match the occupational disease type selected by the user and generate a relevant injury case data set; secondly, the system uses a symbolic regression algorithm to perform a preliminary scoring calculation on each case, models the correlation between different occupational disease characteristics by searching for the optimal injury change expression, automatically explores possible function combinations, and generates basic scoring results; thirdly, the system uses random matrix theory technology to perform statistical analysis on the eigenvalue distribution of the data covariance matrix, identifies outliers and potential patterns, and generates an outlier correction matrix; finally, the system makes precise adjustments to abnormal situations based on the outlier correction matrix to improve the accuracy and robustness of the scoring results and generate a scoring adjustment coefficient.

[0102] In order to solve the accuracy and robustness problems of the scoring model when processing complex data, in some embodiments, the method for performing preliminary scoring calculation based on the relevant injury case data set in step 102 includes: based on the relevant injury case data set, performing data cleaning and normalization processing, identifying features related to injury assessment through feature extraction, and generating a preprocessed injury case data set; based on the preprocessed injury case data set, using a symbolic regression algorithm, performing preliminary scoring calculations on each case, searching for the optimal injury change expression, modeling the relationship between different injury features, and generating a preliminary scoring model for each case; based on the preliminary scoring model for each case, further analyzing the correlation between different injury features, identifying significant correlations to quantify the degree of correlation, and generating a feature correlation analysis report; based on the feature correlation analysis report, optimizing the performance of the scoring model to ensure that the interactions between multiple features are comprehensively considered to generate a basic scoring result.

[0103] In this embodiment, data cleaning refers to removing or correcting erroneous, duplicate, and incomplete records in the database to ensure the quality of the data.

[0104] Normalization is the process of converting data of different dimensions to the same scale for easy comparison and analysis. These processing steps ensure the accuracy of subsequent score calculations.

[0105] The preprocessed injury case data set is a collection of injury case data that has been cleaned and normalized. Features related to injury assessment are identified through feature extraction, and an optimized data set is generated for subsequent scoring calculations.

[0106] The preliminary scoring models for each case are individual scoring models generated after preliminary scoring calculations for each case. These models reflect the basic scoring situation of each case and provide a preliminary scoring framework to help identify which factors have a significant impact on the score.

[0107] The feature correlation analysis report is a document generated by further analyzing the correlation between different injury features. The report quantifies the degree of association between each feature and reveals which features have a significant impact on the scoring results. This report provides an important basis for optimizing the scoring model.

[0108] In the embodiments of the present application, first, data cleaning and normalization processing are performed based on a relevant injury case dataset. By feature extraction, features related to injury assessment are identified to generate a preprocessed injury case dataset. Secondly, based on the preprocessed injury case dataset, the system uses a symbolic regression algorithm to perform a preliminary score calculation for each case. By searching for the optimal injury change expression, the relationship between different injury features is modeled to generate a preliminary score model for each case. Thirdly, based on the preliminary score models of each case, the system further analyzes the correlation between different injury features, identifies significant associations to quantify the degree of association, and generates a feature correlation analysis report. Finally, based on the feature correlation analysis report, the system optimizes the performance of the score model to ensure that the interaction between multiple features is comprehensively considered, and a basic score result is generated.

[0109] The following is a specific example:

[0110] Suppose the user selects a specific type of chronic disease in a medical institution as the evaluation criterion; first, the system performs data cleaning and normalization processing based on a relevant injury case dataset. By feature extraction, features related to chronic disease evaluation, such as disease duration, treatment response, etc., are identified to generate a preprocessed injury case dataset. Secondly, based on the preprocessed injury case dataset, the system uses a symbolic regression algorithm to perform a preliminary score calculation for each case. By searching for the optimal injury change expression, the relationship between different chronic disease features is modeled to generate a preliminary score model for each case. Thirdly, based on the preliminary score models of each case, the system further analyzes the correlation between different chronic disease features, identifies significant associations to quantify the degree of association, and generates a feature correlation analysis report. Finally, based on the feature correlation analysis report, the system optimizes the performance of the score model to ensure that the interaction between multiple features is comprehensively considered, and a basic score result is generated.

[0111] To solve the problems of outliers and potential pattern recognition in the score model and further improve the accuracy and robustness of the score result, in some embodiments, the method for optimization based on the basic score result described in step 102 includes: based on the basic score result, combined with the key features in the original injury case data, new features are introduced or the weights of existing features are adjusted to generate a feature enhanced dataset; based on the feature enhanced dataset, using the random matrix theory technology, the covariance matrix is calculated, and the eigenvalue distribution is statistically analyzed. By the random matrix theory, the non-linear relationship in the data is captured to generate an optimized basic score result; based on the optimized basic score result, the correlation between different injury features is further analyzed, the eigenvalue of the feature value significantly deviating from the normal distribution is identified, and the potential pattern hidden behind the data is identified to generate an outlier and potential pattern recognition report; based on the outlier and potential pattern recognition report, by introducing weights and adjustment factors, the score deviation caused by outliers is corrected to generate an outlier correction matrix.

[0112] In this embodiment, the feature-enhanced dataset is a data set generated by combining the basic scoring results and the key features in the original injury case data. In this process, the system can introduce new features or adjust the weights of existing features to enrich the content of the dataset and ensure the comprehensiveness and accuracy of subsequent analysis.

[0113] The random matrix theory technique is a statistical method used to analyze and identify outliers and potential patterns. Through the statistical analysis of the eigenvalue distribution of the data covariance matrix, it can capture the non-linear relationships in the data and identify outliers and potential patterns.

[0114] The covariance matrix is a matrix that describes the correlations between variables in a multi-dimensional dataset. Each element represents the covariance between two variables, that is, the consistency of their changes.

[0115] The optimized basic scoring result is an improved scoring result generated by statistically analyzing the eigenvalue distribution of the covariance matrix through the random matrix theory technique after introducing new features or adjusting the weights of existing features.

[0116] The outlier and potential pattern recognition report is a document generated by further analyzing the correlations between different injury characteristics. This report quantifies the degree of association between each characteristic, reveals which characteristics deviate significantly from the normal distribution, and identifies the potential patterns hidden behind the data.

[0117] The outlier correction matrix is a set of adjustment factors generated based on the outlier and potential pattern recognition report, used to correct the scoring bias caused by outliers. By introducing weights and adjustment factors, the system can precisely adjust the scoring result to ensure the accuracy and fairness of the final score.

[0118] In the embodiment of this application, first, the system generates a feature-enhanced dataset based on the basic scoring result, combines the key features in the original injury case data, and introduces new features or adjusts the weights of existing features; second, the system calculates the covariance matrix based on the feature-enhanced dataset, conducts a statistical analysis of the eigenvalue distribution, captures the non-linear relationships in the data, and generates an optimized basic scoring result; third, the system further analyzes the correlations between different injury characteristics based on the optimized basic scoring result, identifies the eigenvalue of the characteristic that significantly deviates from the normal distribution and the potential pattern hidden behind the data, and generates an outlier and potential pattern recognition report; finally, the system generates an outlier correction matrix based on the outlier and potential pattern recognition report by introducing weights and adjustment factors to correct the scoring bias caused by outliers.

[0119] The following is a specific example:

[0120] Assume that the user selects a specific type of post-operative recovery as the evaluation criterion in the health insurance claims system; first, based on the basic scoring results, the system combines the key features in the original surgical case data (such as surgery type, post-operative complications, etc.), introduces new features (such as physical indicators during recovery) or adjusts the weights of existing features to generate a feature-enhanced data set; secondly, based on the feature-enhanced data set, the system uses random matrix theory technology to calculate the covariance matrix, performs statistical analysis on the eigenvalue distribution, captures the nonlinear relationship in the data, and generates an optimized basic scoring result; thirdly, based on the optimized basic scoring results, the system further analyzes the correlation between different post-operative recovery features, identifies eigenvalues ​​that significantly deviate from the normal distribution and the potential patterns hidden behind the data, and generates an outlier and potential pattern recognition report; finally, based on the outlier and potential pattern recognition report, the system introduces weights and adjustment factors to correct the scoring deviation caused by outliers and generates an outlier correction matrix.

[0121] In order to solve the flexibility and transparency problems of the scoring model in a complex data environment and further improve the accuracy and interpretability of the scoring results, in some embodiments, the method for optimization based on the scoring adjustment coefficient described in step 103 includes: based on the scoring adjustment coefficient, making fine adjustments, recalibrating various key parameters, and generating an optimized scoring model; based on the optimized scoring model, using a quantum enhancement learning algorithm, using the powerful parallel processing capabilities of quantum computing to accelerate model training, and through the superposition and entanglement characteristics of quantum bits, exploring a large number of possible solution spaces in a short time to generate multiple possible scoring schemes; based on the multiple possible scoring schemes, using causal reasoning technology to perform correlation analysis, explore causal relationships between variables, and construct causal graphs to perform intervention simulation experiments to reveal the independent influence of key factors and generate a detailed causal analysis report; based on the detailed causal analysis report, detailing the specific influence of each key factor to ensure the flexibility and transparency of the scoring system and generate a correction scoring report.

[0122] In this embodiment, the optimized scoring model refers to an improved scoring model generated by finely adjusting the scoring adjustment coefficient and recalibrating various key parameters. This model not only takes into account the original scoring rules, but also integrates new adjustment factors to improve the accuracy and robustness of the scoring.

[0123] Multi-possibility scoring schemes are a series of scoring schemes generated after applying the quantum reinforcement learning algorithm. These schemes provide multiple possible scoring results based on different assumptions and parameter combinations, helping the system to understand the data more comprehensively and select the optimal solution.

[0124] The intervention simulation experiment is based on the causal diagram. By changing the values ​​of certain variables and observing their impact on other variables, the independent influence of key factors is revealed. This method helps to understand the specific impact of different factors on the scoring results and provides transparency and interpretability of the scoring system.

[0125] The Detailed Causality Analysis Report is a document generated after an in-depth analysis of the correlations between different injury characteristics. The report quantifies the causal relationships between the various characteristics, reveals which factors have a significant impact on the scoring results, and provides detailed explanations and instructions.

[0126] The corrected scoring report records in detail the specific impact of each key factor, provides transparency and explanation of the scoring system, and ensures that the scoring results are more accurate and reliable. The report content includes the weight of each factor, independent impact and other relevant information.

[0127] In the embodiment of the present application, first, fine adjustments are made based on the scoring adjustment coefficient, and various key parameters are recalibrated to generate an optimized scoring model; secondly, based on the optimized scoring model, the system uses a quantum enhanced learning algorithm, utilizes the powerful parallel processing capabilities of quantum computing to accelerate model training, and explores a large number of possible solution spaces in a short period of time through the superposition and entanglement characteristics of quantum bits, and generates multiple possible scoring schemes; thirdly, based on the multiple possible scoring schemes, the system uses causal reasoning technology to perform correlation analysis, explores the causal relationship between variables, and constructs a causal graph to perform intervention simulation experiments, revealing the independent influence of key factors, and generating a detailed causal analysis report; finally, based on the detailed causal analysis report, the system describes in detail the specific impact of each key factor, ensures the flexibility and transparency of the scoring system, and generates a correction scoring report.

[0128] Here is a specific example:

[0129] Suppose that a user selects a specific type of occupational disease as an evaluation criterion in a work injury claims system; first, the system makes fine adjustments based on the scoring adjustment coefficient, recalibrates various key parameters such as treatment costs, recovery time, etc., and generates an optimized scoring model; second, based on the optimized scoring model, the system uses a quantum reinforcement learning algorithm, uses the powerful parallel processing capabilities of quantum computing to accelerate model training, and explores a large number of possible solution spaces in a short period of time through the superposition and entanglement characteristics of quantum bits to generate multiple possible scoring schemes; third, based on the multiple possible scoring schemes, the system uses causal reasoning technology to perform correlation analysis, explores the causal relationship between occupational disease characteristics, constructs a causal graph to perform intervention simulation experiments, reveals the independent influence of key factors, and generates a detailed causal analysis report; finally, based on the detailed causal analysis report, the system details the specific impact of each key factor, such as the impact of long-term exposure to hazardous substances on health, ensures the flexibility and transparency of the scoring system, and generates a correction score report.

[0130] In order to solve the stability and consistency problems of the scoring model on different data sets and further improve the model training efficiency and the diversity of scoring results, in some embodiments, the method for accelerated training based on the optimized scoring model in step 103 includes: based on the optimized scoring model, cross-validation and small-scale test set application are used to ensure stability and consistency on different data sets to generate a consistent scoring model; based on the consistent scoring model, a quantum enhanced learning algorithm is used to accelerate model training by utilizing the powerful parallel processing capabilities of quantum computing, and a large number of possible solution spaces are explored in a short time through the superposition and entanglement state characteristics of quantum bits to generate an accelerated training scoring model; based on the accelerated training scoring model, a high-dimensional feature space is constructed, and a variety of injury characteristics and their interrelationships are comprehensively considered to generate a high-dimensional feature representation; based on the high-dimensional feature representation, a variety of potential changes are comprehensively considered, and multiple possible scoring schemes are generated by simulating different condition scenarios.

[0131] The consistent scoring model refers to an improved version that ensures the stability and consistency of the scoring model on different data sets through cross-validation and small-scale test set application.

[0132] The accelerated training scoring model is an improved scoring model generated by using the quantum enhanced learning algorithm to complete a large amount of calculations in a short period of time. This model not only trains faster, but also can more effectively explore complex solution spaces, thereby improving the accuracy and robustness of the scoring results.

[0133] A high-dimensional feature space refers to a data representation that contains multiple dimensions, each dimension representing a feature or attribute.

[0134] High-dimensional feature representation is a data representation generated in a high-dimensional feature space, which can capture more complex relationships and patterns.

[0135] A multi-possibility scoring scheme refers to a series of scoring schemes generated by simulating different condition scenarios and comprehensively considering a variety of potential changes. These schemes are based on different assumptions and parameter combinations to help the system understand the data more comprehensively and select the optimal solution.

[0136] In the embodiments of the present application, firstly, based on the optimized scoring model, cross-validation and small-scale test set applications are used to ensure stability and consistency on different data sets and generate a consistent scoring model; secondly, based on the consistent scoring model, the system uses a quantum enhanced learning algorithm and utilizes the powerful parallel processing capabilities of quantum computing to accelerate model training, and explores a large number of possible solution spaces in a short period of time through the superposition and entanglement characteristics of quantum bits to generate an accelerated training scoring model; thirdly, based on the accelerated training scoring model, the system constructs a high-dimensional feature space, comprehensively considers multiple injury characteristics and their interrelationships, and generates a high-dimensional feature representation; finally, based on the high-dimensional feature representation, the system comprehensively considers multiple potential changes, and generates multiple possible scoring schemes by simulating different condition scenarios.

[0137] Here is a specific example:

[0138] Suppose a user selects a specific type of sports injury as the evaluation criterion in a sports injury claims system; first, based on the optimized scoring model, the system uses cross-validation and small-scale test sets to ensure stability and consistency on different data sets and generate a consistent scoring model; secondly, based on the consistent scoring model, the system uses the quantum reinforcement learning algorithm and the powerful parallel processing capability of quantum computing to accelerate model training. Through the superposition and entanglement characteristics of quantum bits, a large number of possible solution spaces are explored in a short time to generate an accelerated training scoring model; thirdly, based on the accelerated training scoring model, the system constructs a high-dimensional feature space, comprehensively considers multiple sports injury characteristics (such as injured parts, recovery time, etc.) and their relationships, and generates a high-dimensional feature representation; finally, based on the high-dimensional feature representation, the system comprehensively considers multiple potential changes, and generates multiple possible scoring schemes by simulating different condition scenarios (such as different rehabilitation training programs).

[0139] In order to solve the problems of flexibility and user participation of the scoring system in practical applications and further improve the transparency and adaptability of the scoring results, in some embodiments, the method of providing a highly interactive user interface based on the corrected scoring report in step 104 includes: based on the corrected scoring report, providing detailed scoring basis and key factor impact analysis by graphically displaying the corrected scoring results, and generating a highly interactive user interface; based on the highly interactive user interface, allowing users to make flexible adjustments according to actual conditions, collecting user feedback information in real time, and generating a user feedback data set; based on the user feedback data set, the system combines historical data with current user input, uses the Bayesian update method to improve flexibility and adaptability, and generates an adaptive scoring model; based on the adaptive scoring model, deeply integrates all feedback information in the interaction process, ensures the transparency of the scoring system and user participation, and generates an adaptive scoring adjustment plan.

[0140] In this embodiment, the highly interactive user interface is an intuitive and easy-to-operate platform that enables users to view detailed information in the scoring report and flexibly adjust the scoring results according to actual conditions.

[0141] The user feedback dataset is a data set that collects the opinions and suggestions provided by users in real time during the interaction process. This data reflects the user's satisfaction with the rating results, areas that need improvement, etc.

[0142] The Bayesian update method is a statistical method used to update existing probability estimates based on new evidence (such as user feedback). This method allows the system to combine historical data with current user input, dynamically adjust the scoring model parameters, and improve the flexibility and adaptability of the system.

[0143] The adaptive scoring model is a set of scoring rules adjusted based on user feedback and actual needs to ensure that the scoring results are closer to the actual application scenarios. The model can be continuously updated and optimized based on new data and user feedback, improving the accuracy and reliability of the scoring.

[0144] In the embodiments of the present application, firstly, the system generates a highly interactive user interface by graphically displaying the correction scoring results based on the correction scoring report, providing detailed scoring basis and key factor impact analysis; secondly, the system generates a highly interactive user interface, allowing the user to make flexible adjustments according to actual conditions, collect user feedback information in real time, and generate a user feedback data set based on the highly interactive user interface; thirdly, the system generates an adaptive scoring model based on the user feedback data set, combining historical data with current user input, and utilizing the Bayesian update method to improve flexibility and adaptability; finally, the system generates an adaptive scoring model based on the adaptive scoring model, deeply integrating all feedback information in the interactive process, ensuring the transparency of the scoring system and user participation, and generating an adaptive scoring adjustment plan.

[0145] Here is a specific example:

[0146] Suppose that a user selects a specific type of postoperative recovery as an evaluation criterion in a rehabilitation treatment evaluation system; first, based on the correction score report, the system graphically displays the correction score results, provides detailed scoring basis and key factor impact analysis, and generates a highly interactive user interface; secondly, based on the highly interactive user interface, the system allows users to make flexible adjustments according to actual conditions, such as modifying the weights of certain dimensions or adding new scoring rules, collecting user feedback information in real time, and generating a user feedback data set; thirdly, based on the user feedback data set, the system combines historical data with current user input, uses the Bayesian update method to improve flexibility and adaptability, and generates an adaptive scoring model; finally, based on the adaptive scoring model, the system deeply integrates all feedback information in the interactive process to ensure the transparency of the scoring system and user participation, and generates an adaptive scoring adjustment plan.

[0147] This application takes into account that in order to solve the problem of insufficient accuracy and poor robustness of the scoring model in the prior art when processing complex injury data, some embodiments propose a further scoring optimization method. In the prior art, due to the existence of problems such as the influence of different dimensions, feature redundancy, abnormal fluctuations, etc., the scoring results are not accurate enough and difficult to adapt to various application scenarios. Therefore, the embodiment of the invention proposes this optional solution, which aims to solve the above problems and improve the accuracy and robustness of the scoring model through steps such as standardization, dimensionality reduction, preliminary score calculation, exponential smoothing, and the introduction of additional feature fourth power terms. The solution includes:

[0148] Based on the preprocessed injury case data set, a symbolic regression algorithm is used to perform a preliminary score calculation for each case. By searching for the optimal injury change expression, the relationship between different injury characteristics is modeled to generate a preliminary scoring model for each case, including:

[0149] Based on the preprocessed injury case data set, all eigenvalues ​​are converted into a standard normal distribution to eliminate the influence of different dimensions;

[0150] Principal component analysis was used to reduce feature dimensions and retain the maximum amount of information. Pearson correlation coefficient analysis was used to identify and exclude highly redundant and irrelevant features to generate preliminary scores.

[0151] The preliminary score is calculated using the following formula:

[0152]

[0153] Among them, S 1 is the preliminary score; α is the weight parameter used to adjust the overall score scale; β is the contribution of the intensity coefficient to the score; F i is the ith injury characteristic value; μi is the average value of the i-th feature; σ i is the standard deviation of the i-th feature; γ is a constant term to ensure that the lower limit of the score is not zero; η is a shape parameter used to define the form of the gamma function; i is the index of the number of features, from 1 to N; N is the number of features; Γ(η) is the gamma function, which serves as a normalization factor; t is the integral variable;

[0154] Based on the preliminary score, exponential smoothing is applied to reduce the impact of abnormal fluctuations, and by introducing periodic functions and product terms, complex nonlinear relationships between features are captured to enhance the adaptability to actual application scenarios, so as to generate an optimized preliminary score;

[0155] The optimization's preliminary score is calculated using the following formula:

[0156]

[0157] Among them, S 2 is the initial score for optimization; S 1 is the preliminary score; δ is the nonlinear adjustment coefficient to adjust the degree of nonlinear influence; ω is the frequency factor to control the frequency of nonlinear fluctuation; ζ is the influence coefficient of the product term; θ j is the sensitivity parameter of each feature; λ is the weight coefficient of the additional feature; G k is the kth additional eigenvalue; v k is the average value of the kth additional feature; ρ k is the standard deviation of the kth additional feature; k is the index of the additional feature, from 1 to L; L is the number of additional features; F i is the ith injury characteristic value; μ i is the average value of the i-th feature; σ i is the standard deviation of the i-th feature; i is the index of the number of features, from 1 to N; N is the number of features; j is the index of the additional feature, from 1 to M; M is the number of additional features; F j is the jth injury characteristic value; μ j is the average value of the jth feature; σ j is the standard deviation of the jth feature;

[0158] Based on the optimized preliminary score, it is combined with the original feature value and an additional feature quartic term is introduced to further increase the model complexity and generalization expression ability, capture subtle changes in the data, and generate a preliminary scoring model for each case.

[0159] This method aims to use a symbolic regression algorithm combined with statistical analysis to eliminate the influence of different dimensions by standardizing the eigenvalues; using principal component analysis to reduce feature dimensions, retain the maximum amount of information, and exclude highly redundant and insignificant features through the Pearson correlation coefficient; calculating the preliminary score based on the standardized eigenvalues, and then applying exponential smoothing to reduce abnormal fluctuations and capture the complex nonlinear relationship between features; finally, introducing additional feature quartic terms to increase the model complexity and generalization expression ability to capture subtle changes.

[0160] Among them, the preliminary scoring item S 1 Calculated by the preliminary scoring formula; the nonlinear adjustment coefficient δ is determined by experiment; the frequency factor ω is set according to the periodic characteristics; the eigenvalue F i , average value μ i , standard deviation σ i Obtained from the preprocessed injury case data set; the product term influence coefficient ζ is determined by experiment; the sensitivity parameter θ j According to the feature importance setting; additional feature logarithm term Calculated by additional features; additional feature G k , average value v k , standard deviation ρ k Obtained from the additional feature data set; the additional feature weight coefficient λ is set according to actual needs;

[0161] Assume that the user selects a specific type of diabetic complications as the evaluation criterion in a chronic disease management platform; Assume that the weight adjustment term α = 0.8, the strength coefficient β = 1.5, the shape parameter η = 2, the constant term γ = 0.5, the number of features N = 5, and the eigenvalue F i , average value μ i and standard deviation σ i Obtained from the preprocessed injury case dataset;

[0162]

[0163] Assuming the nonlinear adjustment coefficient δ = 0.6, the frequency factor The product term influence coefficient ζ=0.7, the sensitivity parameter θ j = 1.2 (for each feature j), additional feature weight coefficient λ = 0.9, additional feature number M = 3, additional feature number L = 2, additional feature G k , average value v k and standard deviation ρ k Obtained from additional feature datasets;

[0164]

[0165] Assuming the threshold is set to 2.5, since the calculated result 2.8 is greater than the set threshold, it indicates that the score of this diabetes complication case is high, and there may be a higher risk or a more detailed review is required. This is because a higher score reflects the complexity and potential risks of the patient's condition, prompting the medical team to take more stringent monitoring measures and treatment plans to ensure that the patient's health status is effectively managed and improved. Through the above steps, the system not only improves the accuracy and robustness of the scoring model, but also enhances its adaptability under multiple injury assessment criteria, ensuring the reliability and practicality of the scoring results.

[0166] This application takes into account that in order to solve the problems of slow training speed, insufficient feature selection and insufficient robustness of the scoring model in the prior art when processing complex injury data, some embodiments propose a scoring optimization method based on a quantum enhanced learning algorithm. In the prior art, due to the problems of long model training time, feature redundancy, abnormal fluctuations, etc., the scoring results are not accurate enough and difficult to adapt to various application scenarios. Therefore, the embodiment of the invention proposes this optional solution, which aims to solve the above problems through steps such as identifying the correlation between features through covariance matrix analysis, introducing LASSO regression for feature selection, constructing a quantum entangled network, and introducing additional feature quartic terms, and improve the training efficiency, accuracy and robustness of the scoring model. The solution includes:

[0167] Based on the consistency scoring model, the quantum enhanced learning algorithm is used to accelerate model training by taking advantage of the powerful parallel processing capabilities of quantum computing. Through the superposition and entanglement characteristics of quantum bits, a large number of possible solution spaces can be explored in a short period of time to generate a scoring model for accelerated training, including:

[0168] Based on the consistency scoring model, the correlation between features is identified through covariance matrix analysis, LASSO regression is introduced for feature selection, and the key features with the greatest impact are screened out according to the feature importance score to generate a preliminary optimization score based on quantum superposition state;

[0169] The preliminary optimization score based on quantum superposition is calculated using the following formula:

[0170]

[0171] Among them, S quantum is the preliminary optimization score based on quantum superposition state; α′ is the weight parameter used to adjust the overall score scale; β′ is the penalty coefficient to control the degree of influence of feature deviation from the mean; F i is the ith injury characteristic value; μ i is the average value of the i-th feature; γ′ is the influence coefficient of the product term; ω j The frequency factor controls the frequency of the sine function fluctuation; σ jis the standard deviation of the jth feature; i is the index of the feature, from 1 to i; N is the number of features; M is the number of features introduced with nonlinear adjustment; F j is the jth injury characteristic value; μ j is the average value of the jth feature;

[0172] Based on the preliminary optimization score based on the quantum superposition state, a quantum entanglement network is constructed, and exponential smoothing is performed to reduce the influence of abnormal fluctuations by adjusting the quantum bit angle parameter and the frequency factor, and the robustness and adaptability of the model are enhanced by introducing a periodic function to generate a final optimization score based on the quantum entanglement state;

[0173] The final optimization score based on quantum entanglement state is calculated by the following formula:

[0174]

[0175] Among them, T entangled Scoring the final optimization based on quantum entanglement; integral symbol represents the integration of the variable ξ from 0 to 1; δ′ is the influence of the nonlinear adjustment coefficient on the quantum entangled state; θ k is the angle parameter for each quantum bit; G l is the lth additional eigenvalue; ν l is the average value of the lth additional feature; ρ l is the standard deviation of the lth additional feature; k is the index of the quantum bit, from 1 to L; L is the number of quantum bits; l is the feature index of the introduced cubic term, from 1 to P; P is the number of features introduced cubic terms; λ is the influence coefficient of the exponential term; η is the exponential decay coefficient; H m is the mth high-order eigenvalue; ξ m is the average value of the mth high-order feature; τ m is the standard deviation of the mth high-order feature; m is the feature index of the introduced quartic term, from 1 to Q; Q is the number of features introduced with the quartic term; ψ(ξ) is the probability distribution of the system state represented by the quantum state wave function; S quantum Scoring for preliminary optimization based on quantum superposition; F i is the ith injury characteristic value; μ i is the average value of the i-th feature; γ is the product term influence coefficient; ω j The frequency factor controls the frequency of the sine function fluctuation; σ j is the standard deviation of the jth feature;

[0176] Based on the final optimized score based on quantum entangled states, additional characteristic quartic terms are introduced to increase the model complexity and expression ability, and the quantum state wave function is used to represent the probability distribution of the system state, ensuring that the scoring model fully considers all possible quantum states during the training process, thereby generating a scoring model that accelerates training.

[0177] This method aims to use the powerful parallel processing capabilities of quantum computing and the superposition and entanglement characteristics of quantum bits to explore a large number of possible solution spaces in a short period of time. The correlation between features is identified through covariance matrix analysis, and the key features with the greatest impact are screened out by LASSO regression to generate a preliminary optimization score; then, exponential smoothing is performed by constructing a quantum entanglement network to reduce the impact of abnormal fluctuations and capture the complex nonlinear relationship between features; finally, additional feature quartic terms are introduced to increase the model complexity and expression ability, ensuring that the scoring model fully considers all possible quantum states during the training process, and generating a scoring model for accelerated training.

[0178] In the preliminary optimization score based on quantum superposition, the weight adjustment term α ′ : Used to adjust the overall scoring scale to ensure that the scoring results meet actual needs; feature deviation penalty item The exponential function is used to control the influence of the eigenvalue deviating from the mean, emphasizing the importance of the eigenvalue being close to the mean; the product term The sine square function is used to introduce periodic changes, which enhances the adaptability of the scoring model to time series data.

[0179] Among them, the weight adjustment term α ′ Set according to the actual application scenario; penalty coefficient β ′ Determined by experiment; eigenvalue F i , average value μ i and standard deviation σ i Obtained from the preprocessed injury case data set; product term influence coefficient γ ′ According to experience, the frequency factor ω j According to the periodic feature setting; the number of features N and the number of features M introduced with nonlinear adjustment are determined by the total number of features in the data set;

[0180] In the final optimization score based on quantum entanglement, the integral term Indicates the integration of the variable ξ from 0 to 1, and the probability distribution of the system state is represented by the quantum state wave function; the nonlinear adjustment term Capture periodic and nonlinear fluctuations by normalizing eigenvalues ​​with cosine function and cubic function; exponential decay term The high-order features are introduced through the exponential decay function to enhance the stability and discrimination of the scoring model;

[0181] Among them, the nonlinear adjustment coefficient δ ′ Determined by experiment; angle parameter θ k Adjusting the experimental setup by qubit angle; additional feature G l , average value v l and standard deviation ρ l Obtained from an additional feature data set; the number of features P introduced by the cubic term is determined by the total number of features in the data set; the exponential term influence coefficient λ is set according to actual needs; the exponential decay coefficient η is determined by experiments; the high-order feature H m , average value ξ m and standard deviation τ m Obtained from a high-order feature data set; the number of features Q introduced with the quartic term is determined by the total number of features in the data set; the quantum state wave function ψ(ξ) is implemented using a quantum computing library;

[0182] Assume that the user selects a specific type of chronic disease as the evaluation criterion in an elderly health assessment system; the weight adjustment term α ′ =0.9, penalty coefficient β ′ =0.5, product term influence coefficient γ ′ =0.8, frequency factor The number of features N = 5, the number of features introduced with nonlinear adjustment M = 3, the eigenvalue F i , average value μ i and standard deviation σ i Obtained from the preprocessed injury case dataset;

[0183]

[0184] Assuming the nonlinear adjustment coefficient δ ′ =0.7, angle parameter The number of features introduced with the cubic term is P = 4, the exponential term influence coefficient λ = 0.6, the exponential decay coefficient η = 0.3, the number of features introduced with the quartic term is Q = 2, and the additional feature G l , average value v l and standard deviation ρ l Obtained from additional feature dataset; high-order feature H m , average value ξ m and standard deviation τ m Obtained from high-level feature datasets;

[0185]

[0186] Assuming the threshold is set to 3, since the calculated result 3.1 is greater than the set threshold, it indicates that the score of the elderly chronic disease case is high, which may be at higher risk or require more detailed review. This is because a higher score reflects the complexity and potential risks of the patient's condition, prompting the medical team to take more stringent monitoring measures and treatment plans to ensure that the patient's health status is effectively managed and improved. Through the above steps, the system not only improves the training efficiency, accuracy and robustness of the scoring model, but also enhances its adaptive ability under multiple injury assessment standards, ensuring the reliability and practicality of the scoring results.

[0187] Figure 2 A schematic diagram of a structure of an adaptive scoring system applicable to multiple injury assessment standards is provided for the embodiment of the present application. Figure 2 As shown, the device comprises:

[0188] The receiving module 21 is used to receive the injury assessment criteria selected by the user, comprehensively consider the scoring dimensions and weights under different assessment criteria, and generate a customized scoring rule set;

[0189] A retrieval module 22 is used to retrieve relevant injury case data for preliminary scoring based on the customized scoring rule set using a symbolic regression algorithm, optimize the scoring model performance by searching for the optimal injury change expression, capture nonlinear relationships in the data, and use random matrix theory technology to analyze and identify outliers and potential patterns, improve the accuracy and robustness of the scoring results, and generate a scoring adjustment coefficient;

[0190] The analysis module 23 is used to apply the quantum enhanced learning algorithm based on the score adjustment coefficient, use the powerful parallel processing capability of quantum computing to accelerate model training, explore a large number of possible solution spaces, use causal reasoning technology for in-depth analysis, build a causal graph to perform intervention simulation experiments, reveal the independent impact of key factors, and generate a correction score report;

[0191] The generation module 24 is used to provide a highly interactive user interface based on the correction score report, and to generate an adaptive score adjustment plan by combining user feedback to allow the user to make flexible adjustments according to actual conditions.

[0192] Figure 2 The adaptive scoring system applicable to multiple injury assessment criteria can be implemented Figure 1 The implementation principle and technical effect of the adaptive scoring method applicable to multiple injury assessment standards described in the illustrated embodiment will not be described in detail. The specific manner in which each module and unit performs operations in the adaptive scoring system applicable to multiple injury assessment standards in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

[0193] In one possible design, Figure 2 An adaptive scoring system applicable to multiple injury assessment criteria in the illustrated embodiment may be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0194] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0195] The processing component 32 is used to: receive the injury assessment criteria selected by the user, comprehensively consider the scoring dimensions and weights under different assessment criteria, and generate a customized scoring rule set; based on the customized scoring rule set, use the symbolic regression algorithm to retrieve relevant injury case data for preliminary scoring, optimize the scoring model performance by searching for the optimal injury change expression, capture the nonlinear relationship in the data, use random matrix theory technology to analyze and identify outliers and potential patterns, improve the accuracy and robustness of the scoring results, and generate a scoring adjustment coefficient; based on the scoring adjustment coefficient, use the quantum enhancement learning algorithm, use the powerful parallel processing capabilities of quantum computing to accelerate model training, explore a large number of possible solution spaces, use causal reasoning technology for in-depth analysis, build a causal diagram to perform intervention simulation experiments, reveal the independent influence of key factors, and generate a correction score report; based on the correction score report, provide a highly interactive user interface, combine user feedback, allow users to make flexible adjustments according to actual conditions, and generate an adaptive scoring adjustment plan.

[0196] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0197] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component 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 read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0198] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0199] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0200] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0201] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0202] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The illustrated embodiment is an adaptive scoring method applicable to multiple injury assessment criteria.

[0203] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0204] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0205] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0206] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned 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 application.

Claims

1. An adaptive scoring method applicable to multiple injury assessment criteria, characterized in that: include: Receive the injury assessment criteria selected by the user, comprehensively consider the scoring dimensions and weights under different assessment criteria, and generate a customized scoring rule set; Based on the customized scoring rule set, a symbolic regression algorithm is used to retrieve relevant injury case data for preliminary scoring, and the scoring model performance is optimized by searching for the optimal injury change expression, capturing the nonlinear relationship in the data, and using random matrix theory technology to analyze and identify outliers and potential patterns, thereby improving the accuracy and robustness of the scoring results and generating a scoring adjustment coefficient; Based on the score adjustment coefficient, the quantum enhanced learning algorithm is used to accelerate model training by taking advantage of the powerful parallel processing capabilities of quantum computing, explore a large number of possible solution spaces, use causal reasoning technology for in-depth analysis, build a causal graph to perform intervention simulation experiments, reveal the independent impact of key factors, and generate a correction score report; Based on the correction score report, a highly interactive user interface is provided, which combines user feedback to allow the user to make flexible adjustments according to actual conditions and generate an adaptive score adjustment plan.

2. The method according to claim 1, characterized in that Based on the customized scoring rule set, the symbolic regression algorithm is used to retrieve relevant injury case data for preliminary scoring, the scoring model performance is optimized by searching for the optimal injury change expression, the nonlinear relationship in the data is captured, and the random matrix theory technology is used to analyze and identify outliers and potential patterns, improve the accuracy and robustness of the scoring results, and generate the scoring adjustment coefficient, including: Based on the customized scoring rule set, the injury case data in the database is screened and retrieved to match the injury assessment criteria selected by the user, and a relevant injury case data set is generated; Based on the relevant injury case data set, a symbolic regression algorithm is used to perform preliminary scoring calculations on each case, and by searching for the optimal injury change expression, modeling the correlation between different injury characteristics, and automatically exploring possible function combinations, a basic scoring result is generated; Based on the basic scoring results, random matrix theory techniques are used to statistically analyze the distribution of eigenvalues ​​of the data covariance matrix, identify outliers and potential patterns, and generate an outlier correction matrix; Based on the outlier correction matrix, the abnormal situation is accurately adjusted to improve the accuracy and robustness of the scoring results, and the scoring adjustment coefficient is generated.

3. The method according to claim 2, characterized in that Based on the relevant injury case data set, a symbolic regression algorithm is used to perform preliminary scoring calculations on each case, and by searching for the optimal injury change expression, modeling the correlation between different injury characteristics, and automatically exploring possible function combinations, a basic scoring result is generated, including: Based on the relevant injury case data set, data cleaning and normalization processing are performed, and features related to injury assessment are identified through feature extraction to generate a pre-processed injury case data set; Based on the preprocessed injury case data set, a symbolic regression algorithm is used to perform a preliminary score calculation on each case, and the relationship between different injury characteristics is modeled by searching for the optimal injury change expression to generate a preliminary scoring model for each case; Based on the preliminary scoring model of each case, further analyze the correlation between different injury characteristics, identify significant associations to quantify the degree of association, and generate a feature correlation analysis report; Based on the feature correlation analysis report, the scoring model performance is optimized to ensure that the interactions between multiple features are comprehensively considered to generate basic scoring results.

4. The method according to claim 2, characterized in that: Based on the basic scoring results, the random matrix theory technology is used to perform statistical analysis on the eigenvalue distribution of the data covariance matrix, identify outliers and potential patterns, and generate an outlier correction matrix, including: Based on the basic scoring results, combined with key features in the original injury case data, new features are introduced or existing feature weights are adjusted to generate a feature enhanced data set; Based on the feature enhancement data set, the covariance matrix is ​​calculated using random matrix theory technology, and the eigenvalue distribution is statistically analyzed. The nonlinear relationship in the data is captured by random matrix theory to generate an optimized basic scoring result; Based on the optimized basic scoring results, further analyze the correlation between different injury characteristics, identify the characteristic values ​​that significantly deviate from the normal distribution and the potential patterns hidden behind the data, and generate an outlier and potential pattern recognition report; Based on the outlier and potential pattern recognition report, weights and adjustment factors are introduced to correct the scoring deviation caused by outliers and generate an outlier correction matrix.

5. The method according to claim 1, characterized in that Based on the score adjustment coefficient, the quantum enhanced learning algorithm is used to accelerate model training by using the powerful parallel processing capabilities of quantum computing, explore a large number of possible solution spaces, use causal reasoning technology for in-depth analysis, build a causal graph to perform intervention simulation experiments, reveal the independent impact of key factors, and generate a correction score report, including: Based on the score adjustment coefficient, make fine adjustments, recalibrate various key parameters, and generate an optimized score model; Based on the optimized scoring model, the quantum enhanced learning algorithm is used to accelerate model training by taking advantage of the powerful parallel processing capabilities of quantum computing. Through the superposition and entanglement characteristics of quantum bits, a large number of possible solution spaces can be explored in a short period of time to generate multiple possible scoring schemes. Based on the multi-possibility scoring scheme, causal reasoning technology is used to perform correlation analysis, explore the causal relationship between variables, and construct a causal diagram to perform intervention simulation experiments, reveal the independent impact of key factors, and generate a detailed causal relationship analysis report; Based on the detailed causal analysis report, the specific impact of each key factor is explained in detail to ensure the flexibility and transparency of the scoring system and generate a corrected scoring report.

6. The method according to claim 5, characterized in that The scoring model based on the optimization uses the quantum enhanced learning algorithm and the powerful parallel processing capability of quantum computing to accelerate model training. Through the superposition and entanglement characteristics of quantum bits, a large number of possible solution spaces can be explored in a short time to generate multiple possible scoring schemes, including: Based on the optimized scoring model, cross-validation and small-scale test set application are used to ensure stability and consistency on different data sets, and generate a consistent scoring model; Based on the consistency scoring model, the quantum enhanced learning algorithm is used to accelerate model training by taking advantage of the powerful parallel processing capabilities of quantum computing. Through the superposition and entanglement characteristics of quantum bits, a large number of possible solution spaces can be explored in a short period of time to generate a scoring model for accelerated training. Based on the accelerated training scoring model, a high-dimensional feature space is constructed, and a variety of injury characteristics and their interrelationships are comprehensively considered to generate a high-dimensional feature representation; Based on the high-dimensional feature representation, multiple potential changes are comprehensively considered and multiple possible scoring schemes are generated by simulating different condition scenarios.

7. The method according to claim 1, characterized in that Based on the correction score report, a highly interactive user interface is provided, which combines user feedback to allow the user to make flexible adjustments according to actual conditions and generate an adaptive score adjustment plan, including: Based on the correction scoring report, the correction scoring results are graphically displayed, detailed scoring basis and key factor impact analysis are provided, and a highly interactive user interface is generated; Based on the highly interactive user interface, the user is allowed to flexibly adjust according to the actual situation, collect user feedback information in real time, and generate a user feedback data set; Based on the user feedback data set, the system combines historical data with current user input, uses the Bayesian update method to improve flexibility and adaptability, and generates an adaptive scoring model; Based on the adaptive scoring model, all feedback information in the interaction process is deeply integrated to ensure the transparency of the scoring system and user participation, and generate an adaptive scoring adjustment plan.

8. An adaptive scoring system applicable to multiple injury assessment criteria, characterized in that: include: A receiving module is used to receive the injury assessment criteria selected by the user, comprehensively consider the scoring dimensions and weights under different assessment criteria, and generate a customized scoring rule set; A retrieval module is used to retrieve relevant injury case data for preliminary scoring based on the customized scoring rule set and using a symbolic regression algorithm, optimize the scoring model performance by searching for the optimal injury change expression, capture nonlinear relationships in the data, and use random matrix theory technology to analyze and identify outliers and potential patterns, improve the accuracy and robustness of the scoring results, and generate a scoring adjustment coefficient; An analysis module is used to apply a quantum enhanced learning algorithm based on the score adjustment coefficient, use the powerful parallel processing capability of quantum computing to accelerate model training, explore a large number of possible solution spaces, use causal reasoning technology for in-depth analysis, build a causal graph to perform intervention simulation experiments, reveal the independent impact of key factors, and generate a correction score report; The generation module is used to provide a highly interactive user interface based on the correction score report, combine user feedback, allow the user to make flexible adjustments according to actual conditions, and generate an adaptive score adjustment plan.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement an adaptive scoring method applicable to multiple injury assessment standards as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, an adaptive scoring method applicable to multiple injury assessment standards as described in any one of claims 1 to 7 is implemented.

Images