Author(s)

Jiao Tang ¹, Qingkun Yu ²

Affiliation(s)

¹ School of Computer Science and Software Engineering, University of Science and Technology Liaoning, Anshan, 114051, China
² College of Science, University of Science and Technology Liaoning, Anshan, 114051, China

Corresponding Author

Qingkun Yu

ABSTRACT

In the current context, where children's mental health problems are increasingly severe and traditional assessment tools are highly subjective, inefficient, and have poor predictive capabilities, this study focuses on improvement. Data from a specific website is collected and undergoes fine preprocessing. SMOTE is applied to increase the proportion of relevant data from 0.125 to 0.250, enhancing data balance. PCA is used for dimension reduction to identify emotional load and social context as key predictors. After screening multiple regression models, the GBDT model is selected. The R² values of the training and test sets are 0.999 and 0.995, respectively, with the MSE of the training set being 0.250 and the MAE being 0.374, and the MSE of the test set being 0.457 and the MAE being 0.233, showing low prediction errors. The innovation lies in integrating advanced techniques and algorithms to provide a more accurate and reliable assessment method for children's mental health, which is significant for promoting research in this field.

KEYWORDS

Machine Learning, Smote Over-sampling, Principal Component Analysis, Children's Mental Health

CITE THIS PAPER

Jiao Tang, Qingkun Yu, Research on Children's Mental Health Assessment Model Based on Machine Learning. Applied & Educational Psychology (2025) Vol. 6: 161-167. DOI: http://dx.doi.org/10.23977/appep.2025.060123.

REFERENCES

[1] Wang Jiachuang, Dong Longjun. Risk assessment of rockburst using SMOTE oversampling and integration algorithms under GBDT framework [J].Journal of Central South University, 2024, 31(08): 2891-2915.
[2] Zhu Y, Chu Xilin, Le Yao. Differential dimensions of the distribution of dummy words in humanities and social science journal articles and their linguistic functions [J]. International Journal of Chinese Language Teaching, 2025, 6(1).
[3] Wang Peng, Deng Zhi, Fan Yuyuang. Anti-SEU Design for Finite State Machine with One-hot Encoding [J].Telecommunications Technology, 2022, 62(08): 1178-1183.
[4] FU Zhongliang, CHEN Xiaoqing, REN Wei, et al. Stochastic K nearest neighbor algorithm with learning process [J]. Journal of Jilin University (Engineering Edition), 2024, 54(01): 209-220.
[5] Zhu Yuanli, Feng Xiangyang, Yan Qingwu, et al. Study on spatial differentiation of soil organic carbon and main controlling factors in Northeast black soil area farmland based on gradient enhancement decision tree [J]. China Environmental Science, 2024(05): 1-14.

Sponsors, Associates, and Links

---

• Advances in Educational Technology and Psychology
  
  
• Curriculum and Teaching Methodology
  
  
• Transactions on Comparative Education
  
  
• Advances in Vocational and Technical Education
  
  
• Adult and Higher Education