What question did this study set out to answer?

This research aims to develop a model-agnostic framework for screening depression severity among university students using explainable AI techniques.

Model-Agnostic Interpretability for Student Depression Screening: Integrating XAI and Multi-Criteria Model Selection

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates effective depression screening in university students, indicating scalable methodologies are needed.

This research aims to develop a model-agnostic framework for screening depression severity among university students using explainable AI techniques.
Utilized sociodemographic, lifestyle, behavioral, and psychometric characteristics from validated tools in a 27-dimensional feature space.
Trained and tested decision tree, random forest, XGBoost, and LightGBM models for predictive performance and interpretability.
Conducted SHapley Additive exPlanations analysis to identify predictors of severe depression.
Decision Tree and LightGBM achieved the highest classification accuracy at 96.0%.
XGBoost provided the best calibrated probability estimates for depression severity.
The analysis revealed composite psychometric scores as the most significant predictors, showing large effect sizes and high statistical significance.

Rasel et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates effective depression screening in university students, indicating scalable methodologies are needed.

This research aims to develop a model-agnostic framework for screening depression severity among university students using explainable AI techniques.
Utilized sociodemographic, lifestyle, behavioral, and psychometric characteristics from validated tools in a 27-dimensional feature space.
Trained and tested decision tree, random forest, XGBoost, and LightGBM models for predictive performance and interpretability.
Conducted SHapley Additive exPlanations analysis to identify predictors of severe depression.
Decision Tree and LightGBM achieved the highest classification accuracy at 96.0%.
XGBoost provided the best calibrated probability estimates for depression severity.
The analysis revealed composite psychometric scores as the most significant predictors, showing large effect sizes and high statistical significance.

Rasel et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark