WO2022165617A1 - 一种基于行为信息的大学生心理状态评估方法 - Google Patents
一种基于行为信息的大学生心理状态评估方法 Download PDFInfo
- Publication number
- WO2022165617A1 WO2022165617A1 PCT/CN2021/074755 CN2021074755W WO2022165617A1 WO 2022165617 A1 WO2022165617 A1 WO 2022165617A1 CN 2021074755 W CN2021074755 W CN 2021074755W WO 2022165617 A1 WO2022165617 A1 WO 2022165617A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- psychological
- data
- information
- mental state
- mental health
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000003066 decision tree Methods 0.000 claims abstract description 21
- 238000013210 evaluation model Methods 0.000 claims abstract description 16
- 238000012549 training Methods 0.000 claims abstract description 12
- 238000005457 optimization Methods 0.000 claims abstract description 10
- 238000011156 evaluation Methods 0.000 claims abstract description 8
- 230000004630 mental health Effects 0.000 claims description 55
- 230000003542 behavioural effect Effects 0.000 claims description 15
- 208000019901 Anxiety disease Diseases 0.000 claims description 14
- 230000036506 anxiety Effects 0.000 claims description 14
- 230000006996 mental state Effects 0.000 claims description 14
- 238000012360 testing method Methods 0.000 claims description 12
- 238000004891 communication Methods 0.000 claims description 7
- 238000004140 cleaning Methods 0.000 claims description 5
- 230000010354 integration Effects 0.000 claims description 5
- 230000002159 abnormal effect Effects 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 claims description 4
- 238000013480 data collection Methods 0.000 abstract description 7
- 238000004836 empirical method Methods 0.000 abstract description 2
- 230000009323 psychological health Effects 0.000 abstract 4
- 230000006399 behavior Effects 0.000 description 22
- 238000011160 research Methods 0.000 description 14
- 238000004458 analytical method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 235000012054 meals Nutrition 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000010187 selection method Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000000844 transformation Methods 0.000 description 2
- 208000027534 Emotional disease Diseases 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 208000029078 coronary artery disease Diseases 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 208000020016 psychiatric disease Diseases 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/20—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for electronic clinical trials or questionnaires
Definitions
- the invention relates to the technical field of psychological state assessment, in particular to a behavioral information-based psychological assessment method for college students.
- the current research on mental health can be divided into two aspects: traditional mental health research and mental health research based on big data.
- Traditional mental health research infers the characteristics and laws behind the psychological phenomenon based on the explicit expression of mental health.
- the commonly used research methods include observation method, investigation method, test method, experimental method and case method, among which the questionnaire method is mostly used in the experimental method.
- Traditional mental health research is currently relatively mature, and has the advantages of wide research scope, detailed research content, and many research results.
- there are still problems such as limited sample size, time-consuming and labor-intensive data collection and processing, and insufficient timeliness.
- the data of mental health research based on big data mainly comes from wearable devices, mobile smart terminals, Internet behavior records, social activity behavior records, big data storage in cloud computing and data mining, etc.
- the publication date is February 15, 2017, and the Chinese patent document with publication number CN106407623A discloses a mental health assessment method based on an Internet cloud database.
- the RBF neural network assessment model is established through the known samples stored in the cloud database, and the model is used. Applied to the assessment of the mental health of new individuals.
- the patent uses the efficient access capability of data to assess mental health, but it does not include individual data attributes other than personal information;
- the publication date is July 2, 2014, and the Chinese patent document with publication number CN103905486A discloses a mental health state assessment method, which makes full use of the data source of Internet behavior records to establish and train a mental health state based on network behavior characteristics.
- the evaluation model eliminates the influence of individual subjective factors on data collection and improves the accuracy of mental health status evaluation, but its model is established based on the psychological test data filled in on the Internet and the personal information of the fill-in.
- Mental health questionnaires are mapped to mental health status and are models based on static data, so they are lacking in reflecting the comprehensiveness and timeliness of the data.
- the purpose of the present invention is to provide a method for evaluating the mental state of college students based on behavior information.
- a model is established on dynamic time series data, and the behavior of college students is mapped to the state of mental health. It eliminates the influence of individual subjective factors on data collection, improves the accuracy of mental health status assessment, and facilitates subsequent large-scale mental health assessments.
- This paper proposes a new parameter selection method, which is of great significance for improving the accuracy of the model.
- the present invention adopts following technical scheme:
- a method for evaluating the mental state of college students based on behavioral information which is characterized by comprising the steps of: S1. Using a decision tree algorithm, based on students' behavioral information and scores on a mental health questionnaire, respectively establishing and training a mental state evaluation model; S2. Obtaining The behavior information of the new individual is obtained, and the psychological state of the new individual is obtained according to the psychological evaluation model.
- the psychological condition includes depression and anxiety of college students.
- the mental health questionnaire includes: Self-rating Depression Scale, Baker Self-rating Depression Scale, and Self-rating Anxiety Scale, and the score of the questionnaire is calculated according to the results of filling in the questionnaire.
- the student's behavior information includes the following data: the student's basic information, academic performance, card information, library borrowing information and interpersonal communication information.
- step S1 includes the following steps: S1a, performing data integration, using database technology to obtain behavior information of students who have filled out the questionnaires and storing them uniformly; S1b, performing data cleaning, removing abnormal data, removing redundant data, and correcting missing data data to fill;
- step S1 it also continues to include the following steps: S1c, dividing the data into a training set and a test set; S1d, establishing a mental state evaluation model based on a decision tree, and using the differential evolution algorithm for parameter optimization of the decision tree In; S1e, using the training set to train a psychological evaluation model; S1f, using the test set as a new individual, and using the improved evaluation model to obtain a mental health evaluation result.
- a new machine learning parameter tuning method is proposed - using differential evolution algorithm for parameter optimization, which breaks the normal state of using trial and error method, empirical method and subjective judgment for parameter selection in some studies, and proposes a new method.
- the parameter selection method is of great significance for improving the accuracy of the model.
- FIG. 1 is a structural diagram of a decision tree based on a differential evolution algorithm according to an embodiment of the present invention.
- FIG. 2 is a structural diagram of a decision tree based on a differential evolution algorithm according to an embodiment of the present invention.
- FIG. 3 is a flowchart of a differential evolution algorithm according to an embodiment of the present invention.
- a method for psychological evaluation of college students based on behavioral information including steps:
- the psychological test scales include: Self-rating Depression Scale (SDS), Beck Depression Inventory (BDI) and Self-rating Anxiety Scale (Self-rating Anxiety Scale) , SAS).
- SDS Self-rating Depression Scale
- BDI Beck Depression Inventory
- Self-rating Anxiety Scale Self-rating Anxiety Scale
- SAS Self-rating Anxiety Scale
- the behavior information of the student includes the following data: the basic information of the student, the academic record, the one-card information, the library loan information and the interpersonal communication information.
- the basic information of the student includes: gender, age, grade, grade to which the home city belongs, and whether to apply for a poor student.
- the tier of the home city is divided into first-tier cities, second-tier cities, third-tier cities, fourth-tier cities and fifth-tier cities according to the list of Chinese cities.
- the academic achievement includes: whether to obtain a scholarship, the number of subjects taken, the number and proportion of subjects obtained at different levels.
- the one-card information includes: daily consumption and dormitory access control information.
- the dormitory access control data includes: the total number of times the dormitory is opened each month, the times of going out and going back to bed in different time periods on weekdays, and the times of going out and going back to bed in different time periods on weekends and weekends.
- the library borrowing information includes: the total number of times of entering the library since enrollment, the total number of books borrowed since enrollment, and the total number of books borrowed in this semester.
- the interpersonal communication information is mainly obtained according to the frequency of the students swiping their cards to eat with them for three meals a day.
- performing data integration and data cleaning on the basis of existing data includes the following steps:
- Data integration is to use database technology to obtain the behavior information of students who have filled out the questionnaire and store them uniformly;
- Data cleaning includes: removing abnormal data, removing redundant data and filling missing data;
- building a mental health assessment model includes the following steps:
- the parameters that are mainly optimized by the differential evolution algorithm include: the maximum number of features, the minimum number of samples required for internal node subdivision, the minimum number of samples of leaf nodes, the minimum sample weight of leaf nodes, and the minimum impurity of node division .
- the optimization parameters of the differential evolution algorithm include the following steps:
- the optimal parameters of the tree further improve the accuracy of the model.
- FIG. 1 is a schematic flowchart of a mental health assessment method according to an embodiment of the present invention, and the specific steps are as follows:
- Step 101 Obtain the student's psychological status according to the depression self-rating scale, the Baker depression self-rating scale and the anxiety self-rating scale, respectively, and divide the depression and anxiety levels according to the scores of the questionnaire.
- Step 102 obtain the behavior information of 466 students who effectively filled out the questionnaire, mainly including basic academic information, academic performance, one-card information, library borrowing information and interpersonal communication information, the details of these information are shown in Table 3-6,
- the interpersonal communication information is mainly obtained according to the frequency of the students swiping their cards to eat with them for three meals a day.
- the preprocessing includes: data integration, the purpose of which is to obtain the behavior information of students who effectively fill in the questionnaire, and organize it into a wide table for subsequent processing and analysis; data cleaning includes: removing abnormal data , remove redundant data and fill in missing data, the purpose is to improve the quality of the data and make the data better fit the model.
- Step 104 constructing a mental health assessment model of college students based on behavior information.
- the depression level is not 0 is divided into the presence of depression
- the anxiety level is not 0 is divided into the presence of anxiety
- the data set it is divided into training set and test set, with 377 training samples and 89 test samples, and ensure that the proportion of people with depression or anxiety in the test set and training set is equal to the proportion of the total samples;
- FIG. 2 is a structural diagram of a decision tree based on a differential evolution algorithm according to an embodiment of the present invention, a mental health assessment model based on a decision tree is established with a training set as an input, and the differential evolution algorithm is used in the parameter optimization of the decision tree, wherein , the parameter range optimized by the differential evolution algorithm is shown in Table 7, and the termination conditions also consider the maximum number of iterations 1000 times and the mean square error 0.000001.
- other machine learning algorithms may also be used, or other evolutionary algorithms may be used for parameter optimization.
- Step 105 taking the test set as a new individual, using the improved evaluation model to obtain the mental health evaluation result of the new individual, and using the common indicators of machine learning algorithms to measure the model classification result: accuracy rate, recall rate and F1 value.
- the arrival of the era of big data has provided sufficient material for psychological research and created conditions for major breakthroughs in mental health.
- the data-driven mental health analysis of the present invention not only breaks the traditional a priori logic, greatly improves the efficiency and possibility of mental health analysis, but also has a profound impact on psychology in terms of research methods, breaking the traditional mental health Research the current state of "the sample is the population, the individual is the rule, and the situation is the experiment”.
Landscapes
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Primary Health Care (AREA)
- Physics & Mathematics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Epidemiology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
一种基于行为信息的大学生心理状态评估方法，包括步骤：S1、利用决策树算法，基于学生的行为信息和心理健康问卷的得分，分别建立和训练心理状态评估模型；S2、获取新个体的行为信息，根据所述的心理评估模型得到所述新个体的心理状况。上述方法从学生的行为数据出发，提出从行为数据评估学生心理健康状态的方法，消除了个体主观因素对数据收集的影响，提高了心理健康状态评估的准确性，也便于后续进行大规模的心理健康评估，大大提高数据采集的科学性和心理状态评估准确性，同时利用差分进化算法进行参数寻优，打破了现有技术使用试错法、经验法和主观判断进行参数选择的常态。
Description
本发明涉及心理状态评估技术领域，具体涉及一种基于行为信息的大学生心理评估方法。
抑郁症和焦虑症是世界范围内常见的两种心理疾病。据世界卫生组织统计，预计到2020年抑郁症将成为仅次于冠心病的第二大疾病隐患。大学生尤其是大一新生，面对学业压力、环境变化、经济困难和对未来的担忧都有可能在心理健康方面增加患病风险。据2010年大学生杂志社和中国大学生网调查显示，当前我国大学生的心理健康现状不是很乐观，存在过心理方面问题的学生达90％之高，而导致大学生出现心理健康问题的原因主要包括人际交往、就业压力、自我管理能力、情感问题、人生发展、环境不适应和学习压力等方面。当遇到心理问题时，绝大部分学生会选择向朋友、家人倾诉，但仍有38％的同学选择不找任何人，这种讳疾忌医、极力掩饰的行为会导致心理问题长期得不到有效排解，久而久之，负面情绪的淤积可能导致危机心理的出现，严重的还会影响到学生正常的学习和生活，甚至出现极端事件。因此，无论是高校还是心理研究工作者都需要高度重视大学生心理健康。
目前对心理健康的研究可以分为两个方面：传统心理健康研究和基于大数据的心理健康研究。传统心理健康研究根据心理健康的外显表达推测背后的心理现象的特点和规律，常用的研究方法包括观察法、调查法、测验法、实验法和个案法，其中以实验法中的问卷法居多。传统心理健康研 究目前已经比较成熟，具有研究范围广、研究内容细、研究成果多等优势，但还存在样本规模有限、数据采集与处理方面耗时耗力、时效性不足等问题。
基于大数据的心理健康研究的数据主要来源于可穿戴设备、移动智能终端、互联网行为记录、社会活动行为记录、大数据存储于云计算和数据挖掘等。
公开日为2017年2月15日，公开号为CN106407623A的中国专利文献公开了一种基于互联网云端数据库的心理健康评估方法，通过云端数据库存储的已知样本建立RBF神经网络评估模型，并将模型应用于新个体心理健康状况的评估。该专利利用了数据的高效存取能力对心理健康进行评估，但其未包含个体除个人信息以外的数据属性；
公开日为2014年7月2日，公开号为CN103905486A的中国专利文献公开了一种心理健康状态评估方法，充分利用互联网行为记录这一数据来源，建立和训练了基于网络行为特征的心理健康状态评估模型，消除了个体主观因素对数据收集的影响，提高了心理健康状态评估的准确性，但是其模型是根据网络上所填写的心理测试的数据和填写者的个人信息来建立的，是从心理健康问卷映射到心理健康状态，是在静态数据上建立的模型，因此在反映数据的全面性和及时性上有所欠缺。
基于上述，如何充分挖掘学生的行为数据，建立更科学、准确、合理的心理健康评估模型是本领域亟待解决的技术问题。
发明内容
本发明的目的是提供一种基于行为信息的大学生心理状态评估方法，根据大学生的个人信息和日常行为，在动态时序数据上来建立模型，从大 学生的行为映射到心理健康状态。消除了个体主观因素对数据收集的影响，提高了心理健康状态评估的准确性，也便于后续进行大规模的心理健康评估；同时，利用差分进化算法进行参数寻优，打破了部分研究使用试错法、经验法和主观判断进行参数选择的常态，提出了一种新的参数选择方法，对于提高模型准确率有重要意义。
本发明采取以下技术方案：
一种基于行为信息的大学生心理状态评估方法，其特征在于，包括步骤：S1、利用决策树算法，基于学生的行为信息和心理健康问卷的得分，分别建立和训练心理状态评估模型；S2、获取新个体的行为信息，根据所述的心理评估模型得到所述新个体的心理状况。
优选的，所述的心理状况包括大学生抑郁和焦虑两个部分。
优选的，所述心理健康问卷包括：抑郁自评量表、贝克抑郁自评量表、和焦虑自评量表，并根据问卷填写结果计算问卷得分。
优选的，所述学生的行为信息包括以下数据：学生的基本信息、学习成绩、一卡通信息、图书馆借阅信息和人际交往信息。
优选的，步骤S1中，包括以下步骤：S1a、进行数据集成，利用数据库技术获取已填写问卷学生的行为信息并统一存储；S1b、进行数据清洗，清除异常数据、去除冗余数据，并对缺失数据进行填充；
进一步的，步骤S1中，还继续包括下列步骤：S1c、将数据划分为训练集与测试集；S1d、建立基于决策树的心理状态评估模型，并将差分进化算法用于决策树的参数寻优中；S1e、利用所述训练集对心理评估模型进行训练；S1f、将测试集作为新个体，利用改进的评估模型得到心理健康评估结果。
与现有技术相比，本申请具有以下有益效果：
1)充分利用大数据时代带来的数据高效存取能力，从学生的行为数据出发，提出从行为数据评估学生心理健康状态的方法，消除了个体主观因素对数据收集的影响，提高了心理健康状态评估的准确性，也便于后续进行大规模的心理健康评估，大大提高数据采集的科学性和心理状态评估准确性。
2)提出了一种新的机器学习调参方法——利用差分进化算法进行参数寻优，打破了部分研究使用试错法、经验法和主观判断进行参数选择的常态，提出了一种新的参数选择方法，对于提高模型准确率有重要意义。
图1本发明一个实施例的基于差分进化算法的决策树结构图。
图2是本发明一个实施例的基于差分进化算法的决策树结构图。
图3是本发明一个实施例的差分进化算法流程图。
下面结合附图和具体实施例对本发明进一步说明。
首先，对于本发明的方案进行进一步的说明：
一种基于行为信息的大学生心理评估方法，包括步骤：
1)利用改进的决策树算法，基于学生的行为信息和心理健康问卷的得分，分别建立和训练大学生心理健康的评估模型；
2)获取新个体的行为信息，根据所述的心理健康评估模型得到所述新个体的心理健康状况评估结果。
其中，所述心理测试量表包括：抑郁自评量表(Self-rating Depression Scale,SDS)、贝克抑郁自评量表(Beck Depression Inventory,BDI)和 焦虑自评量表(Self-rating Anxiety Scale,SAS)。根据问卷填写结果计算问卷得分，并根据表1和表2划分抑郁和焦虑的等级。
表1：
表2：
其中，所述学生的行为信息包括以下数据：学生的基本信息、学习成绩、一卡通信息、图书馆借阅信息和人际交往信息。
其中，所述学生的基本信息包括：性别、年龄、年级、家乡城市所属等级和是否申请贫困生。
其中，所述家乡城市所属等级根据中国城市分级名单，分为一线城市、二线城市、三线城市、四线城市和五线城市。
其中，所述学习成绩包括：是否获得奖学金、已修科目数、获得不同等级的科目数及比例。
其中，所述一卡通信息包括：日常消费情况和宿舍门禁信息。
其中，所述宿舍门禁数据包括：每月宿舍开门总次数、工作日不同时间段出门和回寝次数和双休日不同时间段出门和回寝次数。
其中，所述图书馆借阅信息包括：入学以来进入图书馆的总次数、入 学以来图书借阅总数和本学期图书借阅总数。
其中，所述人际交往信息主要根据学生每日三餐有学生和其一起刷卡吃饭的频率所得。
其中，所述在已有数据的基础上进行数据集成和数据清洗包括下列步骤：
1、数据集成是利用数据库技术获取已填写问卷学生的行为信息并统一存储；
2、数据清洗包括：清除异常数据、去除冗余数据和填充缺失数据；
其中，构建心理健康评估模型包括下列步骤：
1)将数据划分为训练集与测试集；
2)建立基于决策树的心理健康评估模型，并将差分进化算法用于决策树的参数寻优中；
3)利用所述训练集对心理健康评估模型进行训练；
4)将测试集作为新个体，利用改进的评估模型得到心理健康评估结果，并用准确率、召回率和F1值衡量模型分类结果。
其中，所述步骤2)中，差分进化算法主要优化的参数包括：最大特征数、内部节点再划分所需最小样本数、叶子节点最少样本数、叶子节点最小的样本权重和节点划分最小不纯度。其中差分进化算法优化参数包括下列步骤：
(1)初始化种群，其中，种群规模为100，迭代次数为1000代，缩放因子为0.5，交叉概率为0.6，适应度函数为F1分数(F1分数为分类模型准确率和召回率的调和平均数)；
(2)计算种群中每个个体的适应度值；
(3)判断是否达到终止条件或进化代数达到最大，若满足则输出最优 结果，即决策树的最优参数，若不满足则继续下列步骤；
4)进行变异和交叉操作，得到中间种群；
5)在原种群和中间种群中选择个体，得到新一代种群；
6)进化代数g＝g+1，转步骤(2)。
第二、对于本发明的方案进一步地举例说明：
为便于理解，先简要介绍本发明所依据的科学原理。
在心理测量学中，传统心理健康研究主要根据心理健康的外显表达推测背后的心理现象的特点和规律，从而实现对心理健康的有效解释，预测和干预不同的心理现象，提高人们的生活质量。随着大数据和互联网技术的不断发展，大数据记录了普遍化的人类行为，而个体行为大部分是自然条件发生的，受心理状态支配和影响，所以借助外部观测到的行为数据评估心理健康状态更客观化和合理化。
基于上述原理，根据本发明的一个实施例，从学生的行为信息出发，基于决策数算法建立心理健康评估模型，从抑郁和焦虑两个方面对大学生心理健康进行评估，并利用差分进化算法选择决策树的最优参数，进一步提高模型的准确率。
图1是本发明一个实施例的心理健康评估方法的流程示意图，具体步骤如下：
步骤101，根据抑郁自评量表、贝克抑郁自评量表和焦虑自评量表分别获取学生心理状况，并根据问卷得分划分抑郁和焦虑等级。本实施例共483名学生完成了问卷，且有效问卷数共466份，占96.48％。
步骤102，获取有效填写问卷的466名学生的行为信息，主要包括学术的基本信息、学习成绩、一卡通信息、图书馆借阅信息和人际交往信息， 这些信息的详细内容如表3-6所示，人际交往信息主要根据学生每日三餐有学生和其一起刷卡吃饭的频率所得。
表3
表4
表5
表6
步骤103，一个实施例中，预处理包括：数据集成，其目的是获取有效填写问卷的学生的行为信息，并将其整理到一个宽表中便于后续处理和分析；数据清洗包括：清除异常数据、去除冗余数据和填充缺失数据，其目的是提高数据的质量，使数据更好地适应模型。
步骤104，构建基于行为信息的大学生心理健康评估模型。
考虑到样本数的问题，根据问卷得分将抑郁等级不为0的均划分为存 在抑郁，同理，将焦虑等级不为0的均划分为存在焦虑；
按数据集划分为训练集和测试集，训练样本377个，测试样本89个，并保证在测试集和训练集中抑郁或焦虑的人数占比与总样本占比相等；
图2是本发明一个实施例的基于差分进化算法的决策树结构图，以训练集为输入建立基于决策树的心理健康评估模型，并将差分进化算法用于决策树的参数寻优中，其中，差分进化算法优化的参数范围见表7，终止条件同时考虑最大迭代次数1000次和均方误差0.000001。当然，在其它实施例中，也可以采用其他机器学习算法，或采用其他进化算法进行参数优化。
需要说明的是，基本的决策树算法本身属于现有技术。而将差分进化算法用于决策树的参数寻优中，是本实施例中对决策树算法针对性的改进后的一项利用。
表7
步骤105，将测试集作为新个体，利用改进的评估模型得到新个体的心理健康评估结果，并用机器学习算法常用指标准确率、召回率和F1值衡量模型分类结果。
大数据时代的到来，为心理学研究提供了充分的素材，为心理健康的 重大突破创造了条件。本发明基于数据驱动的心理健康分析不仅打破了传统的先验逻辑，大大提高了心理健康分析的效率和可能性，而且在研究方法上也对心理学产生了深刻的影响，打破了传统心理健康研究“样本即总体、个体即规律、情境即实验”的现有状态。
以上为本发明的可选实施例，本领域普通技术人员还可以在此基础上进行各种变换或改进，在不脱离本发明总的构思的前提下，这些变换或改进都应当属于本发明要求保护的范围之内。
Claims (6)
- 一种基于行为信息的大学生心理状态评估方法，其特征在于，包括步骤：S1、利用改进的决策树算法，基于学生的行为信息和心理健康问卷的得分，分别建立和训练心理状态评估模型；所述改进的决策树算法，是指在基本的决策树算法的基础上，加入了差分进化算法进行参数优化；S2、获取新个体的行为信息，根据所述的心理评估模型得到所述新个体的心理状况。
- 根据权利要求1所述的基于行为信息的大学生心理状态评估方法，其特征在于：所述的心理状况包括大学生抑郁和焦虑两个部分。
- 根据权利要求1所述的基于行为信息的大学生心理状态评估方法，其特征在于：所述心理健康问卷包括：抑郁自评量表、贝克抑郁自评量表、和焦虑自评量表，并根据问卷填写结果计算问卷得分。
- 根据权利要求1所述的基于行为信息的大学生心理状态评估方法，其特征在于：所述学生的行为信息包括以下数据：学生的基本信息、学习成绩、一卡通信息、图书馆借阅信息和人际交往信息。
- 根据权利要求1所述的基于行为信息的大学生心理状态评估方法，其特征在于：步骤S1中，包括以下步骤：S1a、进行数据集成，利用数据库技术获取已填写问卷学生的行为信息并统一存储；S1b、进行数据清洗，清除异常数据、去除冗余数据，并对缺失数据进行填充。
- 根据权利要求5所述的的基于行为信息的大学生心理状态评估方法，其特征在于，步骤S1中，还继续包括下列步骤：S1c、将数据划分为训练集与测试集；S1d、建立基于决策树的心理状态评估模型，并将差分进化算法用于 决策树的参数寻优中；S1e、利用所述训练集对心理评估模型进行训练；S1f、将测试集作为新个体，利用改进的评估模型得到心理健康评估结果。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2021/074755 WO2022165617A1 (zh) | 2021-02-02 | 2021-02-02 | 一种基于行为信息的大学生心理状态评估方法 |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2021/074755 WO2022165617A1 (zh) | 2021-02-02 | 2021-02-02 | 一种基于行为信息的大学生心理状态评估方法 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2022165617A1 true WO2022165617A1 (zh) | 2022-08-11 |
Family
ID=82740654
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2021/074755 WO2022165617A1 (zh) | 2021-02-02 | 2021-02-02 | 一种基于行为信息的大学生心理状态评估方法 |
Country Status (1)
| Country | Link |
|---|---|
| WO (1) | WO2022165617A1 (zh) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115458099A (zh) * | 2022-09-14 | 2022-12-09 | 广州商学院 | 基于问卷测评的人物心理画像获取方法、系统、电子设备 |
| CN115910345A (zh) * | 2022-12-22 | 2023-04-04 | 广东数业智能科技有限公司 | 一种心理健康测评智能预警方法及存储介质 |
| CN117727456A (zh) * | 2023-12-28 | 2024-03-19 | 江苏智慧智能软件科技有限公司 | 一种基于人工智能的产科心理评估模型塑造系统及方法 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002073860A2 (en) * | 2001-03-08 | 2002-09-19 | Adler Richard M | System for analyzing strategic business decisions |
| US20110231356A1 (en) * | 2009-07-01 | 2011-09-22 | Quantum Leap Research, Inc. | FlexSCAPE: Data Driven Hypothesis Testing and Generation System |
| CN108062571A (zh) * | 2017-12-27 | 2018-05-22 | 福州大学 | 基于差分进化随机森林分类器的光伏阵列故障诊断方法 |
| CN108985523A (zh) * | 2018-08-03 | 2018-12-11 | 北京邮电大学 | 一种改进和声搜索优化的小波神经网络短期电力负荷预测方法 |
| CN109036561A (zh) * | 2018-07-10 | 2018-12-18 | 同济大学 | 一种基于行为信息的大学生心理状态评估方法 |
-
2021
- 2021-02-02 WO PCT/CN2021/074755 patent/WO2022165617A1/zh active Application Filing
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002073860A2 (en) * | 2001-03-08 | 2002-09-19 | Adler Richard M | System for analyzing strategic business decisions |
| US20110231356A1 (en) * | 2009-07-01 | 2011-09-22 | Quantum Leap Research, Inc. | FlexSCAPE: Data Driven Hypothesis Testing and Generation System |
| CN108062571A (zh) * | 2017-12-27 | 2018-05-22 | 福州大学 | 基于差分进化随机森林分类器的光伏阵列故障诊断方法 |
| CN109036561A (zh) * | 2018-07-10 | 2018-12-18 | 同济大学 | 一种基于行为信息的大学生心理状态评估方法 |
| CN108985523A (zh) * | 2018-08-03 | 2018-12-11 | 北京邮电大学 | 一种改进和声搜索优化的小波神经网络短期电力负荷预测方法 |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115458099A (zh) * | 2022-09-14 | 2022-12-09 | 广州商学院 | 基于问卷测评的人物心理画像获取方法、系统、电子设备 |
| CN115910345A (zh) * | 2022-12-22 | 2023-04-04 | 广东数业智能科技有限公司 | 一种心理健康测评智能预警方法及存储介质 |
| CN117727456A (zh) * | 2023-12-28 | 2024-03-19 | 江苏智慧智能软件科技有限公司 | 一种基于人工智能的产科心理评估模型塑造系统及方法 |
| CN117727456B (zh) * | 2023-12-28 | 2024-05-24 | 江苏智慧智能软件科技有限公司 | 一种基于人工智能的产科心理评估模型塑造系统及方法 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2022165617A1 (zh) | 一种基于行为信息的大学生心理状态评估方法 | |
| CN109935336B (zh) | 一种儿童呼吸科疾病的智能辅助诊断系统 | |
| Rodgers et al. | Did fertility go up after the Oklahoma City bombing? An analysis of births in metropolitan counties in Oklahoma, 1990–1999 | |
| Hochman et al. | The impact of wealth on subjective well-being: A comparison of three welfare-state regimes | |
| Xu et al. | Influential mechanism of farmers' sense of relative deprivation in the sustainable development of rural tourism | |
| CN109036561A (zh) | 一种基于行为信息的大学生心理状态评估方法 | |
| Majbouri | Against the wind: labor force participation of women and economic instability in Iran | |
| CN111709575A (zh) | 基于c-lstm的学业成绩预测方法 | |
| Gu et al. | Sustainable career development of new urban immigrants: A psychological capital perspective | |
| Lim et al. | Men and women’s different dreams on the future of the gendered division of paid work and household work after COVID-19 in South Korea | |
| Early et al. | The influence of socio‐economic background and gender on school attainment in the United Kingdom: A systematic review | |
| Argys et al. | Losing public health insurance: Tenncare disenrollment and personal financial distress | |
| Kesheng et al. | Data mining and feature analysis of college students’ campus network behavior | |
| Jaffe et al. | Domestic violence and high-conflict divorce: Developing a new generation of research for children. | |
| Zhang | [Retracted] A Novel Deep Learning Model for Analyzing Psychological Stress in College Students | |
| CN113628744A (zh) | 躯体健康量化评估系统及方法 | |
| CN106055875A (zh) | 一种基于大数据的皮纹分析处理装置 | |
| Borenstein et al. | The academic inbreeding controversy: Analysis and evidence from Brazil | |
| Chen et al. | Economic openness, government efficiency, and urbanization | |
| Holthus et al. | Introduction: Making sense of happiness in “unhappy Japan” | |
| Qin et al. | [Retracted] Intervention Algorithm of Horse Racing for Students’ Psychological Disorders Based on Big Data | |
| Binder | Inter-Generational Transmission and the 21st Century Rise in Skilled Mothers' Labor Supply | |
| Dang et al. | Children left behind in China: The role of school fees | |
| CN112951350A (zh) | 一种基于行为信息的大学生心理状态评估方法 | |
| Gelaye et al. | The Effect of Foreign Direct Investment on Ethiopian Economic Growth |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21923648 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 21923648 Country of ref document: EP Kind code of ref document: A1 |