Sports student selection method and platform based on artificial intelligence
By constructing a knowledge graph of athlete abilities and using back-propagation algorithms and time series analysis, the problem of incomplete athlete ability assessment in traditional methods was solved, achieving more accurate athlete selection.
Patent Information
- Application Number
- CN202510662872.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-22
- Publication Date
- 2025-09-16
AI Technical Summary
Traditional sports student selection methods lack a comprehensive analysis of athletes' skill scores and training data, resulting in the neglect of key ability dimensions when predicting athletes' future performance, affecting the fairness and scientific nature of the selection.
Construct a knowledge graph of athlete capabilities, identify node connections and calculate connection strength, use the backpropagation algorithm to set initial weights, combine the time series analysis algorithm to update the selection criteria in real time, and generate a real-time decision support framework.
It improves the efficiency and accuracy of data processing, optimizes the prediction model, ensures the fairness and efficiency of selection, and can more accurately predict athletes' future performance.
Smart Images
Figure CN120655463A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of performance prediction technology, and in particular to an artificial intelligence-based sports student selection method and platform. Background Art
[0002] The field of performance prediction technology involves applying various algorithms and computational models to predict an individual's performance or behavior over a period of time. In sports, performance prediction is particularly crucial, assisting coaching teams in making more informed decisions. By analyzing an athlete's past performance data, physiological parameters, and training load, it can predict an athlete's future development potential, injury risk, and competition results, thereby improving the objectivity and accuracy of the selection process.
[0003] Among them, the AI-based athlete selection method focuses on using AI technology to analyze and predict the future performance of athletes. The key lies in predicting their success rate and development potential in target sports through a comprehensive analysis of their historical performance data and physiological indicators, aiming to make the selection process more scientific, fair, and efficient.
[0004] Traditional methods rely on historical performance data and physiological indicators, lacking comprehensive analysis of athlete skill scores and training data. This leads to the neglect of key ability dimensions when predicting an athlete's future performance. This lack of multidimensional exploration of an athlete's abilities makes it difficult to form a comprehensive evaluation system for an athlete's overall abilities. Without reference to an athlete's training data and skill scores, prediction models can misjudge an athlete's potential and risk, undermining the fairness and scientific nature of selection and ultimately leading to the neglect of potentially outstanding athletes. Summary of the Invention
[0005] The purpose of the present invention is to solve the shortcomings of the existing technology and propose a sports student selection method and platform based on artificial intelligence.
[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solution: a sports student selection method based on artificial intelligence, comprising the following steps: S1: Based on the athletes' competition results, skill scores and training data, organize and summarize them into structured information, perform data cleaning and structuring, build information nodes, and generate the athlete ability knowledge map; S2: Based on the athlete ability knowledge graph, identify and analyze node connections in the graph, calculate the connection strength between nodes, reveal the strength of mutual influence between skills, and map the hidden associations between skills to generate ability correlation analysis results; S3: Based on the results of the ability correlation analysis, a back-propagation algorithm is used to initialize parameters, set initial weights based on the correlation between skills, and automatically adjust parameters to generate a selection and evaluation framework; S4: Based on the selection and evaluation framework, skill weights are updated according to the athletes’ abilities, weight values are automatically modified to reflect the latest ability status, and a weight adjustment strategy is generated; S5: Based on the weight adjustment strategy, a time series analysis algorithm is used to update the selection criteria and weights in real time according to changes in athlete abilities, and the evaluation criteria are adjusted to generate a real-time decision support framework; S6: Based on the real-time decision support framework, continuously monitor the athlete's performance, adjust the parameter settings of the evaluation model according to the data, and generate a selection plan.
[0007] As a further solution of the present invention, the athlete ability knowledge graph includes name, skill rating, and training intensity index; the weight adjustment strategy includes the adjustment range of skill rating weight, the fluctuation range of competition result weight, and the updating frequency of training intensity data; the real-time decision support framework includes athlete competition result analysis, dynamic update of skill rating, and instant feedback on the relationship between training intensity and performance; the selection plan includes comprehensive athlete ranking, skill balance map, and optimization plan for training and competition performance.
[0008] As a further solution of the present invention, based on the athletes' game results, skill scores and training data, the steps of collating and aggregating them into structured information, performing data cleaning and structuring processing, constructing information nodes, and generating an athlete ability knowledge graph are as follows: S101: Based on the athletes' competition results, skill scores and training data, perform data normalization, unify the data format and measurement units, and generate a unified data set; S102: Based on the unified data set, create corresponding information nodes, allocate independent nodes, and assign corresponding identifiers and attribute values to the nodes to generate a node attribute set; S103: Based on the node attribute set, connections are established between nodes, the association between game results, training data, and skill improvement is displayed, and a knowledge graph of athlete capabilities is generated.
[0009] As a further embodiment of the present invention, based on the athlete ability knowledge graph, identifying and analyzing node connections in the graph, calculating the connection strength between nodes, revealing the strength of mutual influence between skills, and mapping hidden associations between skills, the steps of generating ability correlation analysis results are specifically as follows: S201: Based on the athlete ability knowledge graph, identify and record skill nodes and their connections, sort out elements in the graph, and generate a node and connection identification list; S202: Based on the node and connection identification list, weight the connections in the network graph, evaluate the mutual influence between skills, and generate analysis information on the mutual influence between skills; S203: Based on the inter-skill influence analysis information, identify and map hidden associations between skills, analyze clustering of skill nodes, and generate capability correlation analysis results.
[0010] As a further solution of the present invention, based on the results of the ability correlation analysis, a back propagation algorithm is used to initialize parameters, set initial weights according to the correlation between skills, and automatically adjust parameters to generate a selection and evaluation framework. Specifically, the steps are as follows: S301: Based on the capability correlation analysis results, skill data is collected and analyzed, the mutual influence and dependency between skills are evaluated, and initial weights are assigned according to the influence on the skills, the factors are ranked, and a skill parameter initialization table is generated; S302: Based on the skill parameter initialization table, a back propagation algorithm is used to refine each skill parameter, determine the scoring criteria, and establish a scoring rule framework; The back propagation algorithm is based on the formula: ;
[0011] in, is the learning rate, is the error function, For the Layer to The weights of the layers, is the regularization coefficient, is the momentum coefficient, Update the weight value for the previous iteration, is the time attenuation coefficient; S303: Based on the scoring rule framework, combined with the demand and skill evaluation standards, a selection and evaluation process is designed to generate a selection and evaluation framework.
[0012] As a further solution of the present invention, based on the selection and evaluation framework, skill weights are updated according to the athlete's ability display, and the weight values are automatically modified to reflect the latest ability status. The steps of generating a weight adjustment strategy are specifically as follows: S401: Based on the selection and evaluation framework, collect the athlete's performance data in competition and training, and conduct comprehensive collation and summarization to generate an ability demonstration record; S402: Based on the ability demonstration record, comparing the correlation between the athlete's skill indicators and overall performance, and generating a skill weight adjustment basis by adjusting the skill weight; S403: Based on the skill weight adjustment basis, simulate the performance of athletes under differentiated weight configurations, analyze the potential impact of weight adjustment on athlete rankings, and generate a weight adjustment strategy.
[0013] As a further solution of the present invention, based on the weight adjustment strategy, a time series analysis algorithm is used to update the selection criteria and weights in real time in response to changes in athlete abilities, and to adjust the evaluation criteria. The steps for generating a real-time decision support framework are as follows: S501: Based on the weight adjustment strategy, monitor the athlete's performance data in real time during competition and training, analyze the current selection criteria and the athlete's performance, identify the difference between the criteria and the performance, and generate an adjustment need indicator; S502: Based on the adjustment demand indicators, review and update the weights of skill and performance indicators to match the current competitive state and generate refined selection criteria; S503: Based on the refined selection criteria, a time series analysis algorithm is used to adjust the evaluation model, capture changes in athlete capabilities and performance, and generate a real-time decision support framework.
[0014] As a further embodiment of the present invention, based on the real-time decision support framework, the athlete performance is continuously monitored, and the parameter settings of the evaluation model are adjusted according to the data to generate the selection plan. Specifically, the steps are as follows: S601: Based on the real-time decision support framework, classify, summarize and analyze the key indicators of athletes in competition and training to generate athlete performance files; S602: Based on the athlete performance profile, evaluate the adaptability of the model, identify deviations in model parameters, adjust model parameters and structure according to performance data, and generate an evaluation optimization model; S603: Based on the evaluation optimization model, re-analyze the athlete's performance parameters, conduct performance evaluation, determine selection criteria and requirements, capture the athlete's unique abilities and potential, and generate a selection plan.
[0015] An artificial intelligence-based sports student selection platform, the artificial intelligence-based sports student selection platform is used to implement the artificial intelligence-based sports student selection method described above, the platform comprising: The data integration module performs data normalization based on the athletes' competition results, skill scores, and training data, unifies the data format and measurement units, deletes missing values or outliers, and generates a unified data set; The skill graph construction module creates corresponding information nodes based on the unified data set, builds relationships between nodes, displays the correlation between skill improvement and competition results, and generates an athlete ability knowledge graph; The skill interaction analysis module analyzes the connections between nodes based on the athlete ability knowledge graph, calculates the connection strength, reveals the strength of the mutual influence between skills, and obtains the ability correlation analysis results; The evaluation system construction module assigns an initial weight to each skill based on the ability correlation analysis results, automatically adjusts the skill weights and evaluation parameters, and generates a selection and evaluation framework; The skill weight dynamic adjustment module dynamically updates the skill weight based on the selection and evaluation framework, automatically adjusts the weight value, and generates a weight adjustment strategy; The real-time evaluation and adjustment module uses a time series analysis algorithm based on the weight adjustment strategy to monitor changes in athletes' abilities in real time, adjust selection criteria and weights, and obtain a real-time decision support framework; The selection strategy formulation module continuously monitors the performance of athletes based on the real-time decision support framework, adjusts the parameters of the evaluation model, analyzes the comprehensive ranking and skill balance of athletes, and generates a selection plan.
[0016] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, the organization and visualization of data are enhanced by constructing a knowledge graph, so that the performance data and skill scores of athletes are intuitively displayed, which is convenient for coaching teams and selection personnel to quickly grasp the comprehensive ability and potential of athletes. Through the structured processing of athlete data, the efficiency and accuracy of data processing are improved, and the analysis errors caused by low data quality or inconsistent format are reduced. The improvement of the judgment logic, that is, through the comprehensive analysis of structured data and information nodes, the prediction model is optimized, and the prediction of athletes' future performance is more accurate, which provides support for the coaching team to select outstanding athletes and ensures the fairness and efficiency of the selection. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 It is a schematic diagram of the workflow of the present invention; Figure 2 This is a detailed flow chart of S1 of the present invention; Figure 3 This is a detailed flow chart of S2 of the present invention; Figure 4 This is a detailed flow chart of S3 of the present invention; Figure 5 This is a detailed flow chart of S4 of the present invention; Figure 6 This is a detailed flow chart of S5 of the present invention; Figure 7 This is a detailed flow chart of S6 of the present invention; Figure 8 It is a platform flow chart of the present invention. DETAILED DESCRIPTION
[0018] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0019] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.
[0020] Example 1 See also Figure 1 The present invention provides a technical solution: a sports student selection method based on artificial intelligence, comprising the following steps: S1: Based on the athletes' competition results, skill scores and training data, organize and summarize them into structured information, perform data cleaning and structuring, build information nodes, and generate the athlete ability knowledge map; S2: Based on the athlete ability knowledge graph, identify and analyze node connections in the graph, calculate the connection strength between nodes, reveal the strength of mutual influence between skills, and map the hidden connections between skills to generate ability correlation analysis results; S3: Based on the results of the ability correlation analysis, the back-propagation algorithm is used to initialize the parameters. According to the correlation between skills, the initial weights are set and the parameters are automatically adjusted to generate the selection and evaluation framework. S4: Based on the selection and evaluation framework, skill weights are updated according to the athletes’ abilities, weight values are automatically modified, and the latest ability status is reflected to generate weight adjustment strategies; S5: Based on the weight adjustment strategy, a time series analysis algorithm is used to update the selection criteria and weights in real time according to the changes in athletes' abilities, and to adjust the evaluation criteria to generate a real-time decision support framework; S6: Based on the real-time decision support framework, continuously monitor athlete performance, adjust the parameter settings of the evaluation model according to the data, and generate a selection plan.
[0021] The athlete ability knowledge map includes name, skill rating, and training intensity index. The weight adjustment strategy includes the adjustment range of skill rating weight, the fluctuation range of competition result weight, and the update frequency of training intensity data. The real-time decision support framework includes athlete competition result analysis, dynamic update of skill rating, and instant feedback on the relationship between training intensity and performance. The selection plan includes comprehensive athlete ranking, skill balance map, and optimization plan for training and competition performance.
[0022] See also Figure 2 ,Based on the athletes' game results, skill scores and training data,,they are sorted and summarized into structured information, and data,cleaned and structured, and information nodes are constructed to generate,the athlete ability knowledge graph.,Specific steps are as follows: S101: Based on the athletes' competition results, skill scores and training data, perform data normalization, unify the data format and measurement units, and generate a unified data set; In sub-step S101, data normalization is performed based on the athletes' competition results, skill scores, and training data. The minimum-maximum normalization method is used. The MinMaxScaler function in Python is used with the parameter set to feature_range=(0, 1). The data range is adjusted to between 0 and 1, the data format and measurement unit are unified, and a unified data set is generated.
[0023] S102: Based on the unified data set, create corresponding information nodes, allocate independent nodes, and assign corresponding identifiers and attribute values to the nodes to generate a node attribute set; In sub-step S102, based on the unified data set, corresponding information nodes are created. Graph database technology is adopted and the Neo4j platform is used to allocate independent nodes through the CREATE command, and assign corresponding identifiers and attribute values to the nodes. For example, the CREATE (n: Athlete {id: ID, score: Score}) command is used to allocate independent nodes, assign corresponding identifiers and attribute values to the nodes, and generate a node attribute set.
[0024] S103: Based on the node attribute set, establish connections between nodes, display the relationship between game results, training data and skill improvement, and generate an athlete ability knowledge graph.
[0025] In sub-step S103, connections between nodes are established based on the node attribute set. The graph query language Cypher is used to establish relationships between nodes through MATCH and CREATE commands, such as MATCH (a: Athlete), (b: Athlete) WHERE a.score > b.score CREATE (a) - [: BETTER_THAN] -> (b) commands, which display the relationship between competition results, training data and skill improvement, and generate a knowledge graph of athlete capabilities.
[0026] See also Figure 3 Based on the athlete ability knowledge graph, we identify and analyze the node connections in the graph, calculate the connection strength between nodes, reveal the strength of mutual influence between skills, and map the hidden associations between skills. The specific steps to generate the ability correlation analysis results are as follows: S201: Based on the athlete ability knowledge graph, identify and record skill nodes and their connections, sort out the elements in the graph, and generate a node and connection identification list; In sub-step S201, based on the athlete ability knowledge graph, skill nodes and their connections are identified and recorded, and the graph database query language Cypher is used to execute the MATCH command, such as MATCH (n)-[r]->(m)RETURNn, r, m, to sort out the elements in the graph, including nodes and the connections between them, record the node attributes such as ID, skill score and connection type, and generate a node and connection identification list.
[0027] S202: Based on the node and connection identification list, weight the connections in the network graph, evaluate the mutual influence between skills, and generate analysis information on the influence between skills; In sub-step S202, based on the node and connection identification list, the connections in the network graph are weighted. A weighted network analysis method is used to execute the add_weighted_edges_from function through the Python NetworkX library. The parameters include the start point, end point, and weight value of the edge. Weights are assigned to the connections based on the correlation scores between the skills, the mutual influence between the skills is evaluated, and the influence analysis information between the skills is generated.
[0028] S203: Based on the influence analysis information between skills, identify and map the hidden associations between skills, analyze the clustering of skill nodes, and generate capability correlation analysis results.
[0029] In sub-step S203, based on the influence analysis information between skills, the hidden associations between skills are identified and mapped. A community detection algorithm is used, and the community.best_partition function is used through the Python community library. The parameter is a weighted network graph. The clustering of skill nodes is analyzed. According to the connection strength and pattern between skills, the associated skill groups are identified to generate the ability correlation analysis results.
[0030] See also Figure 4 Based on the results of the ability correlation analysis, the back propagation algorithm is used to initialize the parameters, set the initial weights according to the correlation between skills, and automatically adjust the parameters. The specific steps for generating the selection and evaluation framework are as follows: S301: Based on the results of the capability correlation analysis, skill data is collected and analyzed, the mutual influence and dependency between skills are evaluated, and initial weights are assigned based on the influence on the skills. Factors are ranked and a skill parameter initialization table is generated. In sub-step S301, based on the results of the ability correlation analysis, skill data is collected and analyzed, and a linear regression model is adopted. The LinearRegression function in the Python scikit-learn library is used, and the parameters are set to the default values. Based on the mutual influence and dependency evaluation between skills, an initial weight is assigned to each skill, the factors are sorted, and a skill parameter initialization table is generated.
[0031] S302: Based on the skill parameter initialization table, a back propagation algorithm is used to refine each skill parameter, determine the scoring criteria, and establish a scoring rule framework; In sub-step S302, based on the skill parameter initialization table, the backpropagation algorithm is used to build a neural network model using Python's TensorFlow library. The keras.models.Sequential function is used to build the model. The Dense layer is added using the add method to define the model structure. The optimizer is set to Adam, and the loss function selects mean_squared_error. The parameters of each skill are refined, the scoring criteria are determined, and a scoring rule framework is established.
[0032] Back propagation algorithm, according to the formula: ; in, is the learning rate, is the error function, For the Layer to The weights of the layers, is the regularization coefficient, is the momentum coefficient, Update the weight value for the previous iteration, is the time attenuation coefficient; During execution: learning rate Used to control the step size of each parameter update, error function Measure the difference between the predicted value and the true value, weight Defines the strength of each connection in the network, and the newly added regularization coefficient Reduce model complexity by weighting the sum of squares of weights to prevent overfitting, momentum coefficient The concept of the previous step update amount is introduced to move faster along the gradient direction of the objective function. Represents the value of the last weight update, which introduces historical information into the algorithm to enhance stability and speed up convergence, and the time decay coefficient Introduce time variables to adjust the error, match model performance and dynamically adjust requirements over time.
[0033] S303: Based on the scoring rule framework, combined with the demand and skill evaluation standards, design the selection and evaluation process and generate the selection and evaluation framework.
[0034] In sub-step S303, based on the scoring rule framework and combined with the demand and skill evaluation criteria, a selection and evaluation process is designed. Flowchart tools including Microsoft Visio are used to draw steps and decision nodes to indicate each link of the evaluation process and generate a selection and evaluation framework.
[0035] See also Figure 5 Based on the selection and evaluation framework, skill weights are updated according to the athlete's ability display, and the weight values are automatically modified to reflect the latest ability status. The specific steps for generating the weight adjustment strategy are as follows: S401: Based on the selection and assessment framework, collect and summarize the athletes' performance data in competitions and training, and generate performance records; In sub-step S401, based on the selection and evaluation framework, the performance data of athletes in competitions and training are collected, and are comprehensively organized and summarized. The SQL database is used to execute SELECT and INSERT statements, including SELECT performance, training_data FROM athletes INSERT INTO performance_records, to query and store the athlete performance data and generate ability display records.
[0036] S402: Based on the ability demonstration record, compare the correlation between the athlete's skill indicators and overall performance, and generate a skill weight adjustment basis by adjusting the skill weight; In sub-step S402, based on the ability demonstration records, the correlation between the athlete's skill indicators and overall performance is compared. The Pearson correlation coefficient calculation method is adopted. The corr function is used through the Python pandas library with the parameter set to method=pearson to calculate the correlation coefficient between the skill indicators and the overall performance. The skill weight is adjusted according to the correlation result to generate the basis for skill weight adjustment.
[0037] S403: Based on the skill weight adjustment basis, simulate the performance of athletes under differentiated weight configurations, analyze the potential impact of weight adjustment on athlete rankings, and generate a weight adjustment strategy.
[0038] In sub-step S403, based on the skill weight adjustment basis, the performance of athletes under differentiated weight configurations is simulated, and the potential impact of weight adjustment on athlete rankings is analyzed. The Monte Carlo simulation method is used to generate random samples and visualize the results through Python's numpy and matplotlib libraries. Parameters include the skill weight range and the number of simulations. The estimated performance distribution of athletes is simulated to generate a weight adjustment strategy.
[0039] See also Figure 6 Based on the weight adjustment strategy, a time series analysis algorithm is used to update the selection criteria and weights in real time to cope with changes in athlete abilities, and to adjust the evaluation criteria. The specific steps for generating a real-time decision support framework are as follows: S501: Based on the weight adjustment strategy, monitor the athlete's performance data in real time during competition and training, analyze the current selection criteria and athlete performance, identify the difference between the criteria and performance, and generate adjustment need indicators; In sub-step S501, based on the weight adjustment strategy, the performance data of athletes in competitions and training are monitored in real time. The SELECT statement SELECT performance_metrics FROM athlete_performance WHERE date = CURDATE () is executed using the SQL database to analyze the current selection criteria and athlete performance, identify the differences between the criteria and performance, and generate adjustment requirement indicators.
[0040] S502: Based on the adjustment demand indicators, review and update the weights of skill and performance indicators to match the current competitive state and generate refined selection criteria; In sub-step S502, based on the adjustment demand indicators, the weights of the skill and performance indicators are reviewed and updated. The Python pandas library df[weight]=df[performance_metric]*adjustment_factor is used to operate on the data set to match the current competitive status and generate refined selection criteria.
[0041] S503: Based on the refined selection criteria, a time series analysis algorithm is used to adjust the evaluation model, capture changes in athlete ability and performance, and generate a real-time decision support framework.
[0042] In sub-step S503, based on the refined selection criteria, a time series analysis algorithm is used to establish an ARIMA model using the Python statsmodels library. The parameters are set to order = (1, 1, 1) to adjust the evaluation model, capture changes in athletes' abilities and performances, and generate a real-time decision support framework.
[0043] See also Figure 7 Based on the real-time decision support framework, we continuously monitor the athletes’ performance and adjust the parameters of the evaluation model according to the data. The specific steps to generate the selection plan are as follows: S601: Based on a real-time decision support framework, classify, aggregate, and analyze key indicators of athletes in competition and training to generate athlete performance profiles; In sub-step S601, based on the real-time decision support framework, the key indicators of athletes in competition and training are classified, summarized and analyzed. The SQL database is used to execute SELECT statements, including SELECT * FROM athlete_performance WHERE date BETWEEN start_date AND end_date, to perform time series analysis on the indicators and generate athlete performance files.
[0044] S602: Based on the athlete performance profile, evaluate the adaptability of the model, identify deviations in the model parameters, adjust the model parameters and structure according to the performance data, and generate an evaluation optimization model; In sub-step S602, based on the athlete performance profile, the adaptability of the model is evaluated using a logistic regression analysis method. The LogisticRegression function is used through the Python scikit-learn library with the parameter settings including solver=liblinear and multi_class=ovr. The deviations in the model parameters are discovered, and the model parameters and structure are adjusted according to the performance data to generate an evaluation optimization model. S603: Based on the evaluation optimization model, re-analyze the athlete's performance parameters, conduct performance evaluation, determine the selection criteria and requirements, capture the athlete's unique abilities and potential, and generate a selection plan.
[0045] In sub-step S603, based on the evaluation optimization model, the athlete performance parameters are re-analyzed and the support vector machine (SVM) algorithm is used. The SVC function is used through the Python scikit-learn library with the parameter set to kernel=linear to perform performance evaluation, determine the selection criteria and requirements, capture the unique abilities and potential of the athletes, and generate a selection plan.
[0046] See also Figure 8 A sports student selection platform based on artificial intelligence is provided. The sports student selection platform based on artificial intelligence is used to execute the above-mentioned sports student selection method based on artificial intelligence. The platform includes: The data integration module performs data normalization based on the athletes' competition results, skill scores, and training data, unifies the data format and measurement units, deletes missing values or outliers, and generates a unified data set; The skill graph construction module creates corresponding information nodes based on a unified data set, builds relationships between nodes, displays the correlation between skill improvement and competition results, and generates a knowledge graph of athlete capabilities. The skill interaction analysis module analyzes the connections between nodes based on the athlete ability knowledge graph, calculates the connection strength, reveals the strength of the mutual influence between skills, and obtains the ability correlation analysis results; The evaluation system construction module assigns an initial weight to each skill based on the results of the ability correlation analysis, automatically adjusts the skill weights and evaluation parameters, and generates a selection and evaluation framework; The skill weight dynamic adjustment module dynamically updates skill weights based on the selection and evaluation framework, automatically adjusts weight values, and generates weight adjustment strategies; The real-time assessment and adjustment module is based on a weight adjustment strategy and uses a time series analysis algorithm to monitor changes in athlete abilities in real time, adjust selection criteria and weights, and obtain a real-time decision support framework; The selection strategy formulation module is based on a real-time decision support framework, which continuously monitors athlete performance, adjusts the parameters of the evaluation model, analyzes the overall ranking and skill balance of athletes, and generates selection plans.
[0047] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.
Claims
1. A sports student selection method based on artificial intelligence, characterized in that: The following steps are involved: Based on the athletes' competition results, skill scores and training data, the data is organized and summarized into structured information, and data cleaning and structuring are performed to build information nodes and generate athlete ability knowledge graphs; Based on the athlete ability knowledge graph, identifying and analyzing node connections in the graph, calculating the connection strength between nodes, revealing the strength of mutual influence between skills, and mapping hidden associations between skills to generate ability correlation analysis results; Based on the results of the ability correlation analysis, a back-propagation algorithm is used to initialize the parameters, set the initial weights according to the correlation between the skills, and automatically adjust the parameters to generate a selection and evaluation framework; Based on the selection and evaluation framework, skill weights are updated according to the athletes' abilities, weight values are automatically modified, and the latest ability status is reflected to generate a weight adjustment strategy; Based on the weight adjustment strategy, a time series analysis algorithm is used to update the selection criteria and weights in real time according to the changes in athletes' abilities, and to adjust the evaluation criteria to generate a real-time decision support framework; Based on the real-time decision support framework, the athlete performance is continuously monitored, and the parameter settings of the evaluation model are adjusted according to the data to generate a selection plan.
2. The method for selecting athletes based on artificial intelligence according to claim 1, characterized in that: The athlete ability knowledge map includes name, skill rating, and training intensity index. The weight adjustment strategy includes the adjustment range of skill rating weight, the fluctuation range of competition result weight, and the updating frequency of training intensity data. The real-time decision support framework includes athlete competition result analysis, dynamic update of skill rating, and instant feedback on the relationship between training intensity and performance. The selection plan includes comprehensive athlete ranking, skill balance map, and optimization plan for training and competition performance.
3. The method for selecting athletes based on artificial intelligence according to claim 1, characterized in that: Based on the athletes' game results, skill scores, and training data, the following steps are used to organize and summarize them into structured information, clean and structure the data, build information nodes, and generate the athlete ability knowledge graph: Based on the athletes' competition results, skill scores and training data, data normalization is performed to unify the data format and measurement units to generate a unified data set; Based on the unified data set, corresponding information nodes are created, independent nodes are allocated, and corresponding identifiers and attribute values are assigned to the nodes to generate a node attribute set; Based on the node attribute set, connections between nodes are established, the associations between game results, training data, and skill improvement are displayed, and a knowledge graph of athlete capabilities is generated.
4. The method for selecting athletes based on artificial intelligence according to claim 1, characterized in that: Based on the athlete ability knowledge graph, identifying and analyzing node connections in the graph, calculating the connection strength between nodes, revealing the strength of mutual influence between skills, and mapping the hidden associations between skills, the steps for generating ability correlation analysis results are as follows: Based on the athlete ability knowledge graph, identifying and recording skill nodes and their connections, sorting out elements in the graph, and generating a node and connection identification list; Based on the node and connection identification list, weighting the connections in the network graph, evaluating the mutual influence strength between skills, and generating inter-skill influence strength analysis information; Based on the inter-skill influence analysis information, the hidden associations between the skills are identified and mapped, the clustering of the skill nodes is analyzed, and the ability correlation analysis results are generated.
5. The method for selecting athletes based on artificial intelligence according to claim 1, characterized in that: Based on the results of the ability correlation analysis, the back propagation algorithm is used to initialize the parameters, set the initial weights according to the correlation between skills, and automatically adjust the parameters. The specific steps for generating the selection and evaluation framework are as follows: Based on the results of the capability correlation analysis, skill data is collected and analyzed, the mutual influence and dependency between skills are evaluated, and initial weights are assigned according to the influence on the skills, the factors are ranked, and a skill parameter initialization table is generated; Based on the skill parameter initialization table, a back propagation algorithm is used to refine each skill parameter, determine the scoring criteria, and establish a scoring rule framework; The back propagation algorithm is based on the formula: ； in, is the learning rate, is the error function, For the Layer to The weights of the layers, is the regularization coefficient, is the momentum coefficient, Update the weight value for the previous iteration, is the time attenuation coefficient; Based on the scoring rule framework, combined with the demand and skill evaluation standards, a selection and evaluation process is designed to generate a selection and evaluation framework.
6. The method for selecting athletes based on artificial intelligence according to claim 1, characterized in that: Based on the selection and evaluation framework, skill weights are updated according to the athlete's ability display, and the weight values are automatically modified to reflect the latest ability status. The steps for generating a weight adjustment strategy are as follows: Based on the selection and assessment framework, collect and summarize the athletes' performance data in competitions and training, and generate performance records; Based on the ability demonstration records, comparing the correlation between the athlete's skill indicators and overall performance, and generating a skill weight adjustment basis by adjusting the skill weight; Based on the skill weight adjustment basis, simulate the performance of athletes under differentiated weight configurations, analyze the potential impact of weight adjustment on athlete rankings, and generate a weight adjustment strategy.
7. The method for selecting athletes based on artificial intelligence according to claim 1, characterized in that: Based on the weight adjustment strategy, a time series analysis algorithm is used to update the selection criteria and weights in real time in response to changes in athlete abilities, and to adjust the evaluation criteria. The specific steps for generating a real-time decision support framework are as follows: Based on the weight adjustment strategy, real-time monitoring of athlete performance data in competition and training situations is performed, current selection criteria and athlete performance are analyzed, discrepancies between criteria and performance are identified, and adjustment need indicators are generated; Based on the adjustment requirements, review and update the weights of skills and performance indicators to match the current competitive state and generate refined selection criteria; Based on the refined selection criteria, a time series analysis algorithm is used to adjust the evaluation model to capture changes in athletes' abilities and performances and generate a real-time decision support framework.
8. The method for selecting athletes based on artificial intelligence according to claim 1, characterized in that: Based on the real-time decision support framework, the athlete performance is continuously monitored and the parameter settings of the evaluation model are adjusted according to the data. The specific steps for generating the selection plan are as follows: Based on the real-time decision support framework, key indicators of athletes in competition and training are classified, summarized and analyzed to generate athlete performance profiles; Based on the athlete performance profile, evaluating the adaptability of the model, identifying deviations in the model parameters, adjusting the model parameters and structure according to the performance data, and generating an evaluation optimization model; Based on the evaluation optimization model, the athlete's performance parameters are re-analyzed, performance evaluation is conducted, selection criteria and requirements are determined, the unique abilities and potentials of the athletes are captured, and a selection plan is generated.
9. An artificial intelligence-based sports student selection platform, characterized by: According to any one of claims 1 to 8, the method for selecting athletes based on artificial intelligence, the platform comprises: The data integration module performs data normalization based on the athletes' competition results, skill scores, and training data, unifies the data format and measurement units, deletes missing values or outliers, and generates a unified data set; The skill graph construction module creates corresponding information nodes based on the unified data set, builds relationships between nodes, displays the correlation between skill improvement and competition results, and generates an athlete ability knowledge graph; The skill interaction analysis module analyzes the connections between nodes based on the athlete ability knowledge graph, calculates the connection strength, reveals the strength of the mutual influence between skills, and obtains the ability correlation analysis results; The evaluation system construction module assigns an initial weight to each skill based on the ability correlation analysis results, automatically adjusts the skill weights and evaluation parameters, and generates a selection and evaluation framework; The skill weight dynamic adjustment module dynamically updates the skill weight based on the selection and evaluation framework, automatically adjusts the weight value, and generates a weight adjustment strategy; The real-time evaluation and adjustment module uses a time series analysis algorithm based on the weight adjustment strategy to monitor changes in athletes' abilities in real time, adjust selection criteria and weights, and obtain a real-time decision support framework; The selection strategy formulation module continuously monitors the performance of athletes based on the real-time decision support framework, adjusts the parameters of the evaluation model, analyzes the comprehensive ranking and skill balance of athletes, and generates a selection plan.