Scholarlink

<p>This study presents a machine-learning framework for classifying depression severity using behavioral, physiological, and quality-of-life indicators. The methodology integrates Z-score normalization and correlation-based feature grouping to structure heterogeneous variables from the NHANES dataset. A novel Dynamic Monarch Butterfly Optimized Bagging-XGBoost (DMB-Bagging-XGBoost) model is developed to enhance classification stability and maximize predictive performance. Experimental evaluation demonstrates that the proposed model achieves consistently high accuracy across multiple depression severity levels. SHAP-based interpretation identifies key contributing features, enabling transparent assessment of influential behavioral and physiological factors. The findings confirm the capability of the proposed framework to support early identification of individuals at higher risk and to facilitate data-driven recommendations for personalized mental-health interventions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Personalized mental health interventions through ML-based depression and QoL assessment

  • 期刊论文

  • 发表日期：

    2026年1月8日

查看全文
  • K. Saranya,
  • G. Bhuvaneswari

摘要

This study presents a machine-learning framework for classifying depression severity using behavioral, physiological, and quality-of-life indicators. The methodology integrates Z-score normalization and correlation-based feature grouping to structure heterogeneous variables from the NHANES dataset. A novel Dynamic Monarch Butterfly Optimized Bagging-XGBoost (DMB-Bagging-XGBoost) model is developed to enhance classification stability and maximize predictive performance. Experimental evaluation demonstrates that the proposed model achieves consistently high accuracy across multiple depression severity levels. SHAP-based interpretation identifies key contributing features, enabling transparent assessment of influential behavioral and physiological factors. The findings confirm the capability of the proposed framework to support early identification of individuals at higher risk and to facilitate data-driven recommendations for personalized mental-health interventions.