基于机器学习构建贵州省大学生非自杀性自伤行为的预测模型

潘婵; 刘晓容; 石相孜; 赵文欣; 田萌; 陈思远; 张宛筑

doi:10.16835/j.cnki.1000-9817.2023.08.018

基于机器学习构建贵州省大学生非自杀性自伤行为的预测模型

doi: 10.16835/j.cnki.1000-9817.2023.08.018

1.
贵州医科大学公共卫生与健康学院, 贵阳 550025
2.
贵州医科大学医学人文学院

基金项目:

2023年度贵州省教育厅高校人文社会科学研究项目 23GZGXRWJD1931

详细信息

作者简介:
潘婵(1998-), 女, 贵州遵义人, 在读硕士, 主要研究方向为青少年心理健康

通讯作者:
张宛筑, E-mail: 269396139@qq.com

利益冲突声明 所有作者声明无利益冲突。
中图分类号: TP181 B844.2 R193
计量
- 文章访问数: 388
- HTML全文浏览量: 272
- PDF下载量: 65
- 被引次数: 0
出版历程
- 收稿日期: 2023-05-14
- 修回日期: 2023-07-10
- 网络出版日期: 2023-08-26
- 刊出日期: 2023-08-25

A machine learning-based predictive model of non-suicidal self-injurious behavior among college students in Guizhou Province

1.
School of Public Health, Guizhou Medical University, Guiyang(550025), China

摘要

摘要: 目的探索机器学习算法在预测大学生非自杀性自伤(NSSI)行为中的效果, 分析大学生NSSI行为的影响因素, 为促进大学生心理健康提供参考。方法于2022年12月采用分层随机整群抽样方法选取贵州省某高校835名大学生为研究对象, 采用青少年自我伤害行为问卷、家庭功能评定量表、情绪调节自我效能感量表进行施测, 以人口学特征、家庭因素和情绪因素为自变量, 以大学生是否有NSSI行为为因变量, 使用机器学习算法构建预测模型, 包括逻辑回归、支持向量机、决策树、算法梯度提升树、随机森林、AdaBoost。结果大学生NSSI行为检出率为23.23%(194名); NSSI行为组的家庭功能总分、情感交流、自我主义、家庭规则得分高于非NSSI行为组(t值分别为3.02, 3.35, 2.23, 2.87, P值均＜0.05), 非NSSI行为组在情绪调节自我效能感总分、管理消极情绪自我效能感、表达积极情绪自我效能感得分高于NSSI行为组(t值分别为-5.04, -5.48, -2.43, P值均＜0.05)。随机森林、支持向量机、逻辑回归、决策树、算法梯度提升树、AdaBoost的召回率依次为84.3%, 90.6%, 73.4%, 87.5%, 95.3%, 89.0%;F1依次为84.4%, 92.1%, 71.2%, 79.4%, 91.7%, 89.1%;精确度依次为84.4%, 93.5%, 69.1%, 72.7%, 88.4%, 89.1%;AUC依次为0.845, 0.922, 0.706, 0.776, 0.915, 0.891。结论相较于算法梯度提升树、随机森林、逻辑回归和AdaBoost模型, 支持向量机模型预测贵州省大学生是否存在NSSI行为有较好的预测效果。应选择适合模型尽早识别可能存在NSSI行为的学生, 对其进行心理危机干预, 促进学生的心理健康。
- 自我伤害行为 /
- 精神卫生 /
- 预测 /
- 学生
Abstract: Objective To explore the effectiveness of machine learning algorithms in predicting non-suicidal self-injury (NSSI) behavior among college students, and to analyze the influencing factors of NSSI behavior, thus providing a reference for promoting psychological well-being. Methods In December 2022, a stratified random cluster sampling method was used to select 835 college students from a university in Guizhou Province, China. The Adolescent Self-injury Scale, Family Function Assessment Scale, and Emotion Regulation Self-efficacy Scale were used to evaluate the participants. Demographic characteristics, family factors, and emotional factors were taken as independent variables, while the dependent variable was whether college students exhibited NSSI behavior. Machine learning algorithms, including Logistic regression, support vector machine (SVM), decision trees, algorithm gradient boosting trees, random forests, and AdaBoost, were used to construct predictive models. Results The detection rate of NSSI behavior among the college students was 23.23% (194 individuals). The NSSI behavior group scored higher than the non-NSSI behavior group in total family function, emotional communication, egoism, and family rules (t=3.02, 3.35, 2.23, 2.87, P < 0.05). On the other hand, the non-NSSI behavior group scored higher than the NSSI behavior group in total emotion regulation self-efficacy, managing negative emotion self-efficacy, and expressing positive emotion self-efficacy (t=-5.04, -5.48, -2.43, P < 0.05). The recall rates of random forests, SVM, Logistic regression, decision trees, algorithm gradient boosting trees, and AdaBoost were 84.3%, 90.6%, 73.4%, 87.5%, 95.3%, 89.0%, respectively. The F1 scores were 84.4%, 92.1%, 71.2%, 79.4%, 91.7%, 89.1%, respectively. The respective precision rates were 84.4%, 93.5%, 69.1%, 72.7%, 88.4%, 89.1%. The AUC scores were 0.845, 0.922, 0.706, 0.776, 0.915, and 0.891, respectively. Conclusion Compared to the algorithm gradient boosting tree, random forest, Logistic regression, and AdaBoost models, the SVM model has a better predictive effect on whether college students in Guizhou Province exhibits NSSI behavior. It is recommended to use an appropriate model to identify students at risk of NSSI behavior as early as possible and provide psychological crisis interventions to promote their mental health.
- Self-injurious behavior /
- Mental health /
- Forecasting /
- Students
注释:

1) 利益冲突声明 所有作者声明无利益冲突。

HTML全文

图 1 大学生NSSI行为影响因素重要性排序的算法梯度提升树

Figure 1. Gradient boosting tree algorithm for ranking the importance of factors influencing college students' NSSI behavior

下载: 全尺寸图片幻灯片

图 2 大学生NSSI行为影响因素重要性排序的随机森林

Figure 2. Random forests for ranking the importance of factors influencing college students' NSSI behavior

下载: 全尺寸图片幻灯片

图 3 大学生NSSI行为影响因素重要性排序的AdaBoost

Figure 3. AdaBoost for ranking the importance of factors influencing college students' NSSI behavior

下载: 全尺寸图片幻灯片

图 4 大学生NSSI行为影响因素重要性排序的决策树

Figure 4. Decision tree for ranking the importance of factors influencing college students' NSSI behavior

下载: 全尺寸图片幻灯片

表 1 有无NSSI行为的大学生在家庭功能和情绪调节自我效能感不同变量维度及总分间的比较(x±s)

Table 1. Comparison of scores between dimensions of different variables and total scores of FAD and SRESE for college students with and without NSSI behavior(x±s)

有无NSSI行为	人数	家庭功能						情绪调节自我效能感
有无NSSI行为	人数	情感交流	积极沟通	自我主义	问题解决	家庭规则	总分	表达积极情绪自我效能感	管理消极情绪自我效能感	总分
有	194	2.40±0.47	2.75±0.53	2.01±0.50	2.96±0.40	2.45±0.45	2.50±0.22	3.84±0.67	3.16±0.68	3.40±0.57
无	641	2.26±0.52	2.80±0.58	1.91±0.56	2.95±0.50	2.34±0.48	2.43±0.29	3.99±0.73	3.48±0.74	3.66±0.65
t值		3.35	-1.18	2.23	0.06	2.87	3.02	-2.43	-5.48	-5.04
P值		＜0.01	0.24	0.03	0.95	＜0.01	＜0.01	0.02	＜0.01	＜0.01

下载: 导出CSV

表 2 不同机器学习算法预测大学生NSSI行为效果(n=835)

Table 2. Effectiveness of different machine learning algorithms in predicting NSSI behaviors among college students(n=835)

机器学习算法	召回率/%	F1/%	精确度/%	AUC
随机森林	84.3	84.4	84.4	0.845
支持向量机	90.6	92.1	93.5	0.922
逻辑回归	73.4	71.2	69.1	0.706
决策树	87.5	79.4	72.7	0.776
算法梯度提升树	95.3	91.7	88.4	0.915
AdaBoost	89.0	89.1	89.1	0.891

下载: 导出CSV

参考文献(26)

[1]	REN Y, LIN M P, LIU Y H, et al. The mediating role of coping strategy in the association between family functioning and non-suicidal self-injury among Taiwanese adolescents[J]. J Clin Psychol, 2018, 74(7): 1246-1257. doi: 10.1002/jclp.22587
[2]	CASSELS M, WILKINSON P. Non-suicidal self-injury in adolescence[J]. Paediatr Child Health, 2016, 26(12): 554-558. doi: 10.1016/j.paed.2016.08.006
[3]	SWANNELL S, V MARTIN G E, PAGE A, et al. Prevalence of non-suicidal self-injury in nonclinical samples: systematic review, Meta-analysis and Meta-regression[J]. Suic Life Threat Behav, 2014, 44(3): 273-303. doi: 10.1111/sltb.12070
[4]	袁勇, 干敏雷, 尹小俭, 等. 青少年非自杀性自伤行为与抑郁症状的相关性[J]. 中国学校卫生, 2023, 44(5): 659-663. doi: 10.16835/j.cnki.1000-9817.2023.05.005 YUAN Y, GAN M L, YIN X J, et al. Correlation between non-suicidal self-injurious behavior and depressive symptoms in adolescents[J]. Chin J Sch Health, 2023, 44(5): 659-663. (in Chinese) doi: 10.16835/j.cnki.1000-9817.2023.05.005
[5]	SCHATTEN H T, ANDOVER M S, ARMEY M F. The roles of social stress and decision-making in non-suicidal self-injury[J]. Psychiatry Res, 2015, 229(3): 983-991. doi: 10.1016/j.psychres.2015.05.087
[6]	KIEKENS G, HASKING P, BOYES M, et al. The associations between non-suicidal self-injury and first onset suicidal thoughts and behaviors[J]. J Affect Disord, 2018, 239: 171-179. doi: 10.1016/j.jad.2018.06.033
[7]	马鸣, 刘欢, 刘润香. 机器学习在大学生自杀意念预测中的应用[J]. 中国学校卫生, 2022, 43(5): 763-767. doi: 10.16835/j.cnki.1000-9817.2022.05.029 MA M, LIU H, LIU R X. Application of machine learning in the prediction of college studentsc suicidal ideation[J]. Chin J Sch Health, 2022, 43(5): 763-767. (in Chinese) doi: 10.16835/j.cnki.1000-9817.2022.05.029
[8]	徐小明. 机器学习方法用于构建青少年自杀自伤行为预警模型研究[D]. 重庆: 重庆医科大学, 2022. XU X M. Study on the machine learing appoachens using for buding the early wearing model of adolesence with suicidal and self-injurious behaviors[D]. Chongqing: Chongqing Medical University, 2022. (in Chinese)
[9]	AMMERMAN B A, JACOBUCCI R, MCCLOSKEY M S. Using exploratory data mining to identify important correlates of non-suicidal self-injury frequency[J]. Psychol Viol, 2018, 8(4): 515-525. doi: 10.1037/vio0000146
[10]	MICHÉ M, STUDERUS E, MEYER A H, et al. Prospective prediction of suicide attempts in community adolescents and young adults, using regression methods and machine learning[J]. J Affect Dis, 2020, 265: 570-578. doi: 10.1016/j.jad.2019.11.093
[11]	冯玉. 青少年自我伤害行为与个体情绪因素和家庭环境因素的关系[D]. 武汉: 华中师范大学, 2008. FENG Y. The relationship between adolescent self-harming behavior and individual emotional and family environmental factors[D]. Wuhan: Huazhong Normal University, 2008. (in Chinese)
[12]	李荣风, 徐夫真, 纪林芹, 等. 家庭功能评定量表的初步修订[J]. 中国健康心理学杂志, 2013, 21(7): 996-1000. https://www.cnki.com.cn/Article/CJFDTOTAL-JKXL201307018.htm LI R F, XU F Z, JI L Q, et al. Revision of Family Assessment Device[J]. Chin J Health Psychol, 2013, 21(7): 996-1000. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JKXL201307018.htm
[13]	王玉洁, 窦凯, 刘毅. 情绪调节自我效能感量表的修订[J]. 广州大学学报(社会科学版), 2013, 12(1): 45-50. https://www.cnki.com.cn/Article/CJFDTOTAL-GZDX201301007.htm WANG Y J, DOU K, LIU Y. Revision of the Scale Regulatory Emotional Self-efficacy[J]. J Guangzhou Univ(Soc Sci Edit), 2013, 12(1): 45-50. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-GZDX201301007.htm
[14]	熊红星, 张璟, 郑雪. 方法影响结果?方法变异的本质、影响及控制[J]. 心理学探新, 2013, 33(3): 195-199. https://www.cnki.com.cn/Article/CJFDTOTAL-XLXT201303002.htm XIONG H X, ZHANG J, ZHENG X. Methodology influences results? The nature, impact, and control of method variation[J]. Psychol Inquiry, 2013, 33(3): 195-199. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-XLXT201303002.htm
[15]	李磊, 黄水平. 支持向量机原理及其在医学分类中的应用[J]. 中国卫生统计, 2009, 26(1): 22-25. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGWT200901008.htm LI L, HUANG S P. Support vector machine principle and its application in medical classification[J]. Chin J Health Stat, 2009, 26(1): 22-25. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-ZGWT200901008.htm
[16]	BADILLO S, BANFAI B, BIRZELE F, et al. An introduction to machine learning[J]. Clin Pharmacol Therapeut, 2020, 107(4): 871-885.
[17]	CHAWLA N V, BOWYER K W, HALL L O, et al. SMOTE: synthetic minority over-sampling technique[J]. J Artif Intell Res, 2002(16): 321-357.
[18]	顾怀婷, 刘根义, 李晶. 大学生非自杀性自伤行为与抑郁、述情障碍的相关性[J]. 中国健康心理学杂志, 2018, 26(1): 6-9. https://www.cnki.com.cn/Article/CJFDTOTAL-JKXL201801003.htm GU H T, LIU G Y, LI J. Correlation of non-suicidal self-injurious behaviors with depression and narrative disorders in college students[J]. Chin J Health Psychol, 2018, 26(1): 6-9. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JKXL201801003.htm
[19]	潘珍, 毛绍菊, 唐寒梅, 等. 中国大学生非自杀性自伤检出率的Meta分析[J]. 中国学校卫生, 2016, 37(6): 878-881. doi: 10.16835/j.cnki.1000-9817.2016.06.020 PAN Z, MAO S J, TANG H M, et al. Meta-analysis of the detection rate of non-suicidal self-injury among Chinese college students[J]. Chin J Sch Health, 2016, 39(6): 878-881. (in Chinese) doi: 10.16835/j.cnki.1000-9817.2016.06.020
[20]	苟双玉, 杜美杰, 刘晓容, 等. 中国青少年非自杀性自伤检出率的Meta分析[J]. 现代预防医学, 2023, 50(2): 263-271. https://www.cnki.com.cn/Article/CJFDTOTAL-XDYF202302013.htm GOU S Y, DU M J, LIU X R, et al. Meta-analysis of the detection rate of non-suicidal self-injury among Chinese adolescents[J]. Mod Prev Med, 2023, 50(2): 263-271. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-XDYF202302013.htm
[21]	ROTHMAN S, SHER L. Suicide prevention in the COVID-19 era[J]. Prev Med, 2021, 152(Pt 1): 106-547.
[22]	梁楷利, 胡蜀萍, 李玉星, 等. 中小学生非自杀性自伤行为家庭环境因素分析[J]. 现代预防医学, 2021, 48(2): 304-307. https://www.cnki.com.cn/Article/CJFDTOTAL-XDYF202102029.htm LIANG K L, HU S P, LI Y X, et al. Family environment factors of non-suicidal self-injury behaviors among primary and middle school students[J]. Mod Prev Med, 2021, 48(2): 304-307. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-XDYF202102029.htm
[23]	张桂梅, 向婷, 蔡艺娴, 等. 青少年非自杀性自伤患者父母人格特征及认知调节情绪策略[J]. 中国学校卫生, 2022, 43(10): 1525-1528. doi: 10.16835/j.cnki.1000-9817.2022.10.020 ZHANG G M, XIANG T, CAI Y X, et al. Personality and cognitive regulation emotion strategies among parents of adolescents with non-suicidal self-injury[J]. Chin J Sch Health, 2022, 43(10): 1525-1528. (in Chinese) doi: 10.16835/j.cnki.1000-9817.2022.10.020
[24]	RAJULA H S R, VERLATO G, MANCHIA M, et al. Comparison of conventional statistical methods with machine learning in medicine: diagnosis, drug development, and treatment[J]. Medicina(Kaunas), 2020, 56(9): 455.
[25]	LINTHICUM P K, SCHAFER K M, RIBEIRO J D. Machine learning in suicide science: applications and ethics[J]. Behav Sci Law, 2019, 37(3): 214-222.
[26]	GRENDAS L N, CHIAPELLA L, RODANTE D E, et al. Comparison of traditional model-based statistical methods with machine learning for the prediction of suicide behaviour[J]. J Psychiatr Res, 2022, 145: 85-91.