A psychological crisis intelligent early warning system based on psychological evaluation of college students

Abstract

This invention provides an intelligent early warning system for psychological crises based on psychological assessment of college students, relating to the field of data processing technology, including: defining an evaluation function to assess the performance of the current risk assessment model; and randomly generating a set of risk assessment model parameters as the current solution. x current At the current temperature T Next, proceed L The next iteration; in each iteration, the current solution... x current Apply random perturbation to generate new solutions x new ; Calculate the evaluation function value of the new solution f ( x new The evaluation function value of the current solution. f ( x current );if f ( x new )< f ( x current If the solution is incorrect, then accept the new solution and update. x current = x new ;if f ( x new )≥ f ( x current If ), then accept the new solution, update the temperature, if T < T f If the condition is met, the iteration stops; otherwise, it continues. After optimization using the simulated annealing algorithm, the final configuration parameters are output. Based on the final configuration parameters, the corresponding risk assessment model is configured to obtain the final risk assessment model. This invention can improve the accuracy of early warning information generation.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to an intelligent early warning system for psychological crises based on psychological assessments of college students. Background Technology

[0002] The construction of a psychological crisis early warning system is particularly important. Traditional technical solutions have some shortcomings in practical applications and may not be able to fully meet the needs of universities for monitoring and early warning of students' mental health. The following are the main shortcomings of traditional technical solutions: Traditional methods of collecting psychological assessment data are sometimes limited to specific scales and questionnaires. These data sources are relatively singular and sometimes fail to fully reflect the students' true psychological state. In addition, the lack of collection of multi-dimensional data such as students' daily behavior and social interactions limits the accuracy and comprehensiveness of psychological assessments.

[0003] In traditional technical solutions, risk assessment models are sometimes configured based on experience or fixed parameter settings. Such configurations are often ill-suited to the varying psychological characteristics of different student groups and may fail to accurately identify high-risk individuals with psychological crises. Due to limitations in data collection and risk assessment models, early warning information generated by traditional technical solutions often results in false alarms or omissions. This inaccurate information not only affects the decision-making effectiveness of school administrators and psychological counselors but also... Summary of the Invention

[0004] This invention provides an intelligent early warning system for psychological crises based on psychological assessment of college students, which can improve the accuracy of early warning information generation.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: Firstly, a psychological crisis intelligent early warning system based on college student psychological assessment includes: The data collection module is used to collect psychological assessment data from university students; The data analysis module, connected to the data collection module, is used to set the initial temperature, temperature drop rate, and termination temperature of the simulated annealing algorithm. T f and the number of iterations at each temperature L Define an evaluation function to assess the performance of the current risk assessment model; randomly generate a set of risk assessment model parameters as the current solution. x current At the current temperature T Next, proceed L The next iteration; in each iteration, the current solution... x current Apply random perturbation to generate a new solution xnew; calculate the evaluation function value of the new solution. f (x new The evaluation function value of the current solution. f ( x current );if f ( x new )< f ( x current If the solution is incorrect, then accept the new solution and update. x current = x new ;if f ( x new )≥ f ( x current If ), then accept the new solution, update the temperature, if T < T f If the condition is met, the iteration stops; otherwise, the iteration continues. After optimization by the simulated annealing algorithm, the final configuration parameters are output. The corresponding risk assessment model is configured according to the final configuration parameters to obtain the final risk assessment model. The early warning generation module, connected to the data analysis module, is used to predict high-risk individuals with psychological crises based on the final risk assessment model; and to generate corresponding early warning information based on the high-risk individuals with psychological crises, the early warning information including risk level and recommended measures. The early warning release module, connected to the early warning generation module, is used to release the generated early warning information to predetermined recipients, including school administrators, psychological counselors, and students' parents.

[0006] Furthermore, psychological assessment data of university students will be collected, including: Define the optional range of questionnaire distribution time and the frequency range of psychological counseling record collection; determine the method of collecting daily behavior data; Define a fitness function to evaluate the effectiveness of each data collection strategy; randomly generate a series of data collection strategy combinations, each combination including questionnaire distribution time, consultation record collection frequency, and daily behavior data collection methods; simulate each strategy combination and use the fitness function to evaluate the effectiveness of each strategy; perform crossover and mutation operations on the selected strategies to generate new strategy combinations, evaluate and select the newly generated strategies, iterate for multiple rounds until the fitness reaches a preset threshold to obtain the corresponding strategy as the final data collection strategy; Based on the final data collection strategy, actual data collection will be conducted to obtain the final mental health questionnaire answers, psychological counseling records, and daily behavior data.

[0007] Furthermore, the selected strategies are subjected to crossover and mutation operations to generate new strategy combinations. These newly generated strategies are then evaluated and selected, iterating multiple rounds until the fitness reaches a preset threshold, yielding the corresponding strategy as the final data collection strategy. This includes: Randomly select two policies from the currently selected policy set as parents, randomly select an intersection point, split the two parent policies at the intersection point, and swap the latter half to generate two new policy combinations. For newly generated strategy combinations, the parameters are randomly changed. A fitness function is used to evaluate the new strategy combination, comparing its fitness with that of the strategies in the current strategy set. If the fitness of the new strategy is higher than that of the strategies in the current strategy set, the new strategy is used to replace the existing strategy; otherwise, the strategies in the current strategy set are retained, and iterative updates are performed to continue the next round of crossover, mutation, and selection. If the preset number of iterations is reached, iteration stops, and the specific parameters of the current corresponding strategy are output, including questionnaire distribution time, consultation record collection frequency, and daily behavior data collection methods. The specific parameters of the current corresponding strategy are used as the final data collection strategy.

[0008] Furthermore, the psychological assessment data includes answers to mental health questionnaires, psychological counseling records, and daily behavior data.

[0009] Furthermore, the formula for calculating the evaluation function is as follows: ; in, Indicates the total number of samples; The total number of categories indicates the number of output categories for the model in a classification task; Indicates category The weights; Indicates that the i-th sample belongs to category The real labels on it; Indicates that the i-th sample belongs to category The predicted probability; Represents the regularization coefficient; It is the model's first One parameter; This represents the weighting coefficient of the entropy term.

[0010] Furthermore, the formula for calculating the fitness function is as follows: ; in, This represents the total number of all questionnaire distribution timing strategies combined. It represents the number of valid responses received under the i-th strategy combination; It represents the total number of questionnaires distributed under the i-th strategy combination; It is the number of all combinations of psychological counseling frequency strategies; It is the first The number of valid psychological counseling records collected under each strategy combination; It is the first Record the total frequency of collection under each strategy combination; It is the number of all combinations of behavioral data collection strategies; It is the number of valid daily behavior data collected under the kth strategy combination; It is the total amount of behavioral data collected under the k-th strategy combination; , and Indicates the weighting coefficient; , and Indicates the index value.

[0011] Furthermore, the recipients include school administrators, psychological counselors, and students' parents.

[0012] Secondly, a psychological crisis intelligent early warning method based on college student psychological assessment includes: Collect psychological assessment data from university students; Set the initial temperature, temperature drop rate, and termination temperature for the simulated annealing algorithm. T f and the number of iterations at each temperature L Define an evaluation function to assess the performance of the current risk assessment model; randomly generate a set of risk assessment model parameters as the current solution. x current At the current temperature T Next, proceed L The next iteration; in each iteration, the current solution... x current Apply random perturbation to generate a new solution xnew; calculate the evaluation function value of the new solution. f ( x new The evaluation function value of the current solution. f ( x current );if f ( x new )< f ( x current If the solution is incorrect, then accept the new solution and update. x current = x new ;if f ( x new )≥ f (x current If ), then accept the new solution, update the temperature, if T < T f If the condition is met, the iteration stops; otherwise, the iteration continues. After optimization by the simulated annealing algorithm, the final configuration parameters are output. The corresponding risk assessment model is configured according to the final configuration parameters to obtain the final risk assessment model. The final risk assessment model predicts high-risk individuals for psychological crisis; based on these high-risk individuals, corresponding early warning information is generated, including risk level and recommended measures. The generated warning information will be sent to designated recipients, including school administrators, psychological counselors, and students' parents.

[0013] Thirdly, a computing device, comprising: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0014] Fourthly, a computer-readable storage medium storing a program that is executed by a processor using the method.

[0015] The above-described solution of the present invention has at least the following beneficial effects: By introducing a simulated annealing algorithm to optimize the configuration parameters of the risk assessment model, the system can more accurately identify high-risk individuals with psychological crises. This optimization method improves the accuracy of early warnings, reduces false alarms and missed alarms, and enables school administrators, psychological counselors, and parents to intervene and provide assistance more effectively.

[0016] The system can conduct personalized assessments based on students' specific psychological test data using a risk assessment model. This personalized assessment method can better reflect each student's unique psychological state and provide each student with more accurate early warnings of psychological crises.

[0017] Through automated data collection, analysis, and early warning generation processes, the system can provide early warning information in real time or near real time. This immediacy helps school administrators, psychological counselors, and parents to identify problems in a timely manner and take appropriate measures to intervene, thereby preventing the psychological crisis from worsening.

[0018] The system can promptly release early warning information to school administrators, psychological counselors, and students' parents, promoting information sharing and collaborative work among multiple parties. This collaborative mechanism helps to form a comprehensive and multi-level psychological crisis response system, improving the efficiency and effectiveness of psychological crisis management. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of a psychological crisis intelligent early warning system based on college student psychological assessment provided by an embodiment of the present invention. Detailed Implementation

[0020] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0021] like Figure 1 As shown, an embodiment of the present invention proposes a psychological crisis intelligent early warning system based on college student psychological assessment, comprising: The data collection module is used to collect psychological assessment data of college students. The recipients include school administrators, psychological counselors, and students' parents. The data analysis module, connected to the data collection module, is used to set the initial temperature, temperature drop rate, and termination temperature of the simulated annealing algorithm. T f and the number of iterations at each temperature L Define an evaluation function to assess the performance of the current risk assessment model; randomly generate a set of risk assessment model parameters as the current solution. x current At the current temperature T Next, proceed L The next iteration; in each iteration, the current solution... x current Apply random perturbation to generate a new solution xnew; calculate the evaluation function value of the new solution. f ( x new The evaluation function value of the current solution. f ( x current );if f ( x new )< f ( x current If the solution is incorrect, then accept the new solution and update. x current = xnew ;if f ( x new )≥ f ( x current If ), then accept the new solution, update the temperature, if T < T f If the condition is met, the iteration stops; otherwise, the iteration continues. After optimization by the simulated annealing algorithm, the final configuration parameters are output. The corresponding risk assessment model is configured according to the final configuration parameters to obtain the final risk assessment model. The early warning generation module, connected to the data analysis module, is used to predict high-risk individuals with psychological crises based on the final risk assessment model; and to generate corresponding early warning information based on the high-risk individuals with psychological crises, the early warning information including risk level and recommended measures. The early warning release module, connected to the early warning generation module, is used to release the generated early warning information to predetermined recipients, including school administrators, psychological counselors, and students' parents.

[0022] In this embodiment of the invention, by introducing a simulated annealing algorithm to optimize the configuration parameters of the risk assessment model, the system can more accurately identify high-risk individuals with psychological crises. This optimization method improves the accuracy of early warnings, reduces false alarms and missed alarms, and enables school administrators, psychological counselors, and parents to intervene and provide assistance more effectively. The system can conduct personalized assessments based on students' specific psychological test data using the risk assessment model. This personalized assessment method better reflects each student's unique psychological state and provides more accurate early warnings of psychological crises for each student. Through automated data collection, analysis, and early warning generation processes, the system can provide early warning information in real-time or near real-time. This immediacy helps school administrators, psychological counselors, and parents to promptly identify problems and take appropriate intervention measures, thereby preventing further deterioration of the psychological crisis. The system can promptly release early warning information to school administrators, psychological counselors, and parents, promoting information sharing and collaborative work among multiple parties. This collaborative mechanism helps to form a comprehensive, multi-layered psychological crisis response system, improving the efficiency and effectiveness of psychological crisis management.

[0023] In this embodiment of the invention, collecting psychological assessment data from university students includes: Define the optional range of questionnaire distribution time and the frequency range of psychological counseling record collection; determine the method of collecting daily behavior data; Define a fitness function to evaluate the effectiveness of each data collection strategy; randomly generate a series of data collection strategy combinations, each combination including questionnaire distribution time, consultation record collection frequency, and daily behavior data collection methods; simulate each strategy combination and use the fitness function to evaluate the effectiveness of each strategy; perform crossover and mutation operations on the selected strategies to generate new strategy combinations, evaluate and select the newly generated strategies, iterate for multiple rounds until the fitness reaches a preset threshold to obtain the corresponding strategy as the final data collection strategy; Based on the final data collection strategy, actual data collection will be conducted to obtain the final mental health questionnaire answers, psychological counseling records, and daily behavior data.

[0024] In this embodiment of the invention, by defining a fitness function and randomly generating and iteratively optimizing combinations of data collection strategies, the method can find the most suitable data collection method for a specific group of college students. This strategy optimization ensures the targeting and efficiency of data collection, thereby improving the accuracy and validity of psychological assessment data. This method integrates mental health questionnaires, psychological counseling records, and daily behavior data to comprehensively capture students' psychological state from multiple perspectives. This multi-dimensional data collection helps to gain a deeper understanding of students' mental health, providing a solid foundation for subsequent psychological crisis early warning and intervention. Through multiple rounds of strategy iteration and optimization, collection methods more likely to provide high-quality data can be selected, ensuring that the final collected psychological assessment data has higher reliability and validity. Optimized data collection strategies ensure the timeliness and continuity of data, which is crucial for the real-time response of psychological crisis early warning systems. Timely data updates can help relevant personnel discover potential psychological problems earlier and take effective intervention measures. Through refined data collection, schools and educators can more accurately understand the psychological state and needs of each student, thereby providing them with more personalized care and support. Optimized data collection strategies can help reduce unnecessary data collection work and improve resource utilization efficiency. This can not only reduce the workload of relevant staff, but also reduce the cost of data collection.

[0025] In this embodiment of the invention, crossover and mutation operations are performed on the selected strategy to generate new strategy combinations. The newly generated strategies are evaluated and selected, and the process is iterated for multiple rounds until the fitness reaches a preset threshold to obtain the corresponding strategy as the final data collection strategy. This includes: Randomly select two policies from the currently selected policy set as parents, randomly select an intersection point, split the two parent policies at the intersection point, and swap the latter half to generate two new policy combinations. For newly generated strategy combinations, the parameters are randomly changed. A fitness function is used to evaluate the new strategy combination, comparing its fitness with that of the strategies in the current strategy set. If the fitness of the new strategy is higher than that of the strategies in the current strategy set, the new strategy is used; otherwise, the strategies in the current strategy set are retained, and iterative updates are performed to continue the next round of crossover, mutation, and selection. If a preset number of iterations is reached, iteration stops, and the specific parameters of the current corresponding strategy are output, including questionnaire distribution time, consultation record collection frequency, and daily behavior data collection method. The specific parameters of the current corresponding strategy are used as the final data collection strategy. The psychological assessment data includes answers to mental health questionnaires, psychological consultation records, and daily behavior data.

[0026] In this embodiment of the invention, crossover and mutation operations can generate new and diverse combinations of data collection strategies. This diversity helps explore a broader strategy space, making it more likely to find the globally optimal data collection strategy. This method can adaptively adjust and optimize the data collection strategy based on feedback from the fitness function. Through continuous iteration and selection, the system can gradually approach the data collection strategy best suited to the current environment and needs. By setting a fitness threshold and the number of iteration rounds, this method can quickly converge to a satisfactory data collection strategy with limited computational resources, improving the efficiency and effectiveness of the optimization process. Because this method is based on the idea of ​​a genetic algorithm, it has good flexibility and scalability, allowing for easy adjustment of parameters such as the fitness function and crossover / mutation rules to adapt to different data collection needs and scenarios. The optimized data collection strategy can more effectively collect high-quality psychological assessment data, which not only reduces invalid and redundant data collection work but also ensures that the collected data is more representative and reliable. The optimized data collection strategy can provide strong data support for school administrators, psychological counselors, etc., helping them to more accurately understand students' mental health status and thus formulate more scientific decisions and intervention measures.

[0027] In this embodiment of the invention, the calculation formula for the evaluation function is as follows: ; in, Indicates the total number of samples; The total number of categories indicates the number of output categories for the model in a classification task; Indicates category The weights; Indicates that the i-th sample belongs to category The real labels on it; Indicates that the i-th sample belongs to category The predicted probability; Represents the regularization coefficient; It is the model's first One parameter; This represents the weighting coefficient of the entropy term.

[0028] In this embodiment of the invention, the evaluation function comprehensively considers classification accuracy, model complexity, and the uncertainty of prediction results. Through a combination of weighted cross-entropy loss, regularization, and entropy terms, it can comprehensively evaluate the model's performance and avoid biases that may arise from a single metric. This is achieved by introducing class weights. This allows for flexible adjustment of the importance of different classes in the evaluation function, which is particularly useful for handling imbalanced datasets. The model's focus on these classes can be increased by increasing the weight of the minority class. (Regularization term in the evaluation function) This helps prevent overfitting by penalizing the sum of squared model parameters, encouraging the model to use fewer parameters and thus improving its generalization ability. Entropy term The introduction of entropy encourages models to produce more certain predictions. High entropy values ​​imply high uncertainty in predictions; by minimizing the entropy term, the model is incentivized to make more confident predictions, thus improving its robustness. The design of this evaluation function makes it applicable not only to specific datasets or models but also highly scalable and generalizable. By adjusting the weight coefficients, it can be adapted to different task requirements and model characteristics.

[0029] In this embodiment of the invention, the formula for calculating the fitness function is: ; in, This represents the total number of all questionnaire distribution timing strategies combined. It represents the number of valid responses received under the i-th strategy combination; It represents the total number of questionnaires distributed under the i-th strategy combination; It is the number of all combinations of psychological counseling frequency strategies; It is the first The number of valid psychological counseling records collected under each strategy combination; It is the first Record the total frequency of collection under each strategy combination; It is the number of all combinations of behavioral data collection strategies; It is the number of valid daily behavior data collected under the kth strategy combination; It is the total amount of behavioral data collected under the k-th strategy combination; , and Indicates the weighting coefficient; , and Indicates the index value.

[0030] In this embodiment of the invention, the fitness function comprehensively considers the effects of three dimensions: questionnaire distribution, psychological counseling record collection, and daily behavior data collection. This ensures a comprehensive evaluation of the data collection strategy. This multi-dimensional evaluation method helps to identify the strengths and weaknesses of the strategy in different aspects, thereby enabling targeted optimization. By quantifying the effective response rate, effective counseling record rate, and effective behavior data collection rate under each strategy combination, the fitness function provides specific numerical indicators for the effectiveness of the strategy, making the comparison between different strategies more objective and accurate. This is achieved by introducing weighting coefficients. , and The fitness function allows for flexible adjustment of the importance of each dimension, enabling it to be customized to meet different needs and preferences and better adapt to various assessment scenarios. Beyond evaluating strategy effectiveness, the fitness function also serves as a guide for strategy optimization. By comparing the fitness values ​​of different strategies, it's possible to identify which strategies perform better in which aspects, thus guiding subsequent strategy adjustments and optimizations. Using fitness functions to evaluate and optimize strategies allows for more effective identification of optimal data collection strategies, improving data collection efficiency and accuracy and providing stronger data support for subsequent psychological crisis early warning and intervention. The quantitative assessment results provided by the fitness function can offer decision support to school administrators, psychological counselors, and others. They can select the optimal data collection strategy based on the fitness value, ensuring the accuracy and validity of psychological assessment data.

[0031] A psychological crisis intelligent early warning method based on college student psychological assessment includes: Collect psychological assessment data from university students; Set the initial temperature, temperature drop rate, and termination temperature for the simulated annealing algorithm. T f and the number of iterations at each temperature L Define an evaluation function to assess the performance of the current risk assessment model; randomly generate a set of risk assessment model parameters as the current solution. x current At the current temperature T Next, proceed L The next iteration; in each iteration, the current solution... x current Apply random perturbation to generate a new solution xnew; calculate the evaluation function value of the new solution. f ( x new The evaluation function value of the current solution. f ( x current );if f (x new )< f ( x current If the solution is incorrect, then accept the new solution and update. x current = x new ;if f ( x new )≥ f ( x current If ), then accept the new solution, update the temperature, if T < T f If the condition is met, the iteration stops; otherwise, the iteration continues. After optimization by the simulated annealing algorithm, the final configuration parameters are output. The corresponding risk assessment model is configured according to the final configuration parameters to obtain the final risk assessment model. The final risk assessment model predicts high-risk individuals for psychological crisis; based on these high-risk individuals, corresponding early warning information is generated, including risk level and recommended measures. The generated warning information will be sent to designated recipients, including school administrators, psychological counselors, and students' parents.

[0032] A computing device, comprising: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0033] A computer-readable storage medium storing a program that is executed by a processor using the method.

[0034] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A psychological crisis intelligent early warning system based on college student psychological assessment, characterized in that, include: The data collection module is used to collect psychological assessment data from university students; The data analysis module, connected to the data collection module, is used to set the initial temperature, temperature drop rate, and termination temperature of the simulated annealing algorithm. T f and the number of iterations at each temperature L Define an evaluation function to assess the performance of the current risk assessment model; randomly generate a set of risk assessment model parameters as the current solution. x current At the current temperature T Next, proceed L The next iteration; In each iteration, for the current solution x current Apply random perturbation to generate new solutions x new ; Calculate the evaluation function value of the new solution. f ( x new The evaluation function value of the current solution. f ( x current ); if f ( x new )< f ( x current If the solution is incorrect, then accept the new solution and update. x current = x new ;if f ( x new )≥ f ( x current If ), then accept the new solution, update the temperature, if T < T f If the condition is met, stop iterating; otherwise, continue iterating. After optimization using the simulated annealing algorithm, the final configuration parameters are output. Configure the corresponding risk assessment model based on the final configuration parameters to obtain the final risk assessment model; The early warning generation module, connected to the data analysis module, is used to predict high-risk individuals with psychological crises based on the final risk assessment model. Based on high-risk individuals with psychological crises, corresponding early warning information is generated, including risk level and recommended measures; The early warning release module, connected to the early warning generation module, is used to release the generated early warning information to predetermined recipients, including school administrators, psychological counselors, and students' parents.

2. The intelligent early warning system for psychological crises based on college student psychological assessment as described in claim 1, characterized in that, Collect psychological assessment data from university students, including: Define the optional range of questionnaire distribution time and the frequency range of psychological counseling record collection; determine the method of collecting daily behavior data; Define a fitness function to evaluate the effectiveness of each data collection strategy; randomly generate a series of data collection strategy combinations, each combination including questionnaire distribution time, consultation record collection frequency, and daily behavior data collection methods; simulate each strategy combination and use the fitness function to evaluate the effectiveness of each strategy; perform crossover and mutation operations on the selected strategies to generate new strategy combinations, evaluate and select the newly generated strategies, iterate for multiple rounds until the fitness reaches a preset threshold to obtain the corresponding strategy as the final data collection strategy; Based on the final data collection strategy, actual data collection will be conducted to obtain the final mental health questionnaire answers, psychological counseling records, and daily behavior data.

3. The intelligent early warning system for psychological crises based on college student psychological assessment as described in claim 2, characterized in that, The selected strategy is subjected to crossover and mutation operations to generate new strategy combinations. These newly generated strategies are then evaluated and selected. This process is repeated multiple times until the fitness reaches a preset threshold, yielding the corresponding strategy as the final data collection strategy. This includes: Randomly select two policies from the currently selected policy set as parents, randomly select an intersection point, split the two parent policies at the intersection point, and swap the latter half to generate two new policy combinations. For newly generated strategy combinations, the parameters are randomly changed. A fitness function is used to evaluate the new strategy combination, comparing its fitness with that of the strategies in the current strategy set. If the fitness of the new strategy is higher than that of the strategies in the current strategy set, the new strategy is used to replace the existing strategy; otherwise, the strategies in the current strategy set are retained, and iterative updates are performed to continue the next round of crossover, mutation, and selection. If the preset number of iterations is reached, iteration stops, and the specific parameters of the current corresponding strategy are output, including questionnaire distribution time, consultation record collection frequency, and daily behavior data collection methods. The specific parameters of the current corresponding strategy are used as the final data collection strategy.

4. The intelligent early warning system for psychological crises based on college student psychological assessment as described in claim 3, characterized in that, The psychological assessment data includes answers to mental health questionnaires, psychological counseling records, and daily behavior data.

5. The intelligent early warning system for psychological crises based on college student psychological assessment as described in claim 4, characterized in that, The formula for calculating the evaluation function is: ； in, Indicates the total number of samples; The total number of categories indicates the number of output categories for the model in a classification task; Indicates category The weights; Indicates that the i-th sample belongs to category The real labels on it; This represents the predicted probability of the i-th sample in the class; Represents the regularization coefficient; It is the model's first One parameter; This represents the weighting coefficient of the entropy term.

6. The intelligent early warning system for psychological crises based on college student psychological assessment as described in claim 5, characterized in that, The formula for calculating the fitness function is: ； in, This represents the total number of all questionnaire distribution timing strategies combined. It represents the number of valid responses received under the i-th strategy combination; It represents the total number of questionnaires distributed under the i-th strategy combination; It is the number of all combinations of psychological counseling frequency strategies; It is the first The number of valid psychological counseling records collected under each strategy combination; It is the first Record the total frequency of collection under each strategy combination; It is the number of all combinations of behavioral data collection strategies; It is the number of valid daily behavior data collected under the kth strategy combination; It is the total amount of behavioral data collected under the k-th strategy combination; , and Indicates the weighting coefficient; , and Indicates the index value.

7. The intelligent early warning system for psychological crises based on college student psychological assessment as described in claim 6, characterized in that, The recipients include school administrators, psychological counselors, and students' parents.

8. A psychological crisis intelligent early warning method based on college student psychological assessment, characterized in that, The system is used to perform the method as described in any one of claims 1 to 7, comprising: Collect psychological assessment data from university students; Set the initial temperature, temperature drop rate, and termination temperature for the simulated annealing algorithm. T f and the number of iterations at each temperature L Define an evaluation function to assess the performance of the current risk assessment model; randomly generate a set of risk assessment model parameters as the current solution. x current At the current temperature T Next, proceed L The next iteration; in each iteration, the current solution... x current Apply random perturbation to generate a new solution xnew; calculate the evaluation function value of the new solution. f ( x new The evaluation function value of the current solution. f ( x current );if f ( x new )< f ( x current If the solution is incorrect, then accept the new solution and update. x current = x new ;if f ( x new )≥ f ( x current If ), then accept the new solution, update the temperature, if T < T f If the condition is met, the iteration stops; otherwise, the iteration continues. After optimization by the simulated annealing algorithm, the final configuration parameters are output. The corresponding risk assessment model is configured according to the final configuration parameters to obtain the final risk assessment model. The final risk assessment model predicts high-risk individuals for psychological crisis; based on these high-risk individuals, corresponding early warning information is generated, including risk level and recommended measures. The generated warning information will be sent to designated recipients, including school administrators, psychological counselors, and students' parents.

9. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to perform the method as described in claim 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the method as described in claim 8.

Images