Psychological Risk Profiling for Post-COVID-19 Anxiety Using Interpretable Ensemble Learning

DOI

10.2991/978-94-6239-664-7_25How to use a DOI?

Keywords

Covid-19; Machine learning; Voting ensemble; SHAP; anxiety; mental health

Abstract

COVID-19 pandemic has left significant physical and mental health consequences with anxiety disorders being particularly prevalent among recovered patients. This study presents a machine learning framework to predict post-COVID-19 anxiety using data from 1,000 recovered individuals in Bangladesh. Several models including SVM, XGBoost, LightGBM, CNN and a heterogeneous voting ensemble were developed and evaluated. Data preprocessing involved z-score normalization, label encoding and the SMOTE to address class imbalance. For interpretability, SHAP was employed to identify key risk factors influencing predictions. The proposed heterogeneous voting ensemble, combining Random Forest, Gradient Boosting, SVM, Logistic Regression and K-Nearest Neighbors with soft voting achieved superior performance compared to individual models and prior studies. It attained 98% accuracy with MAE of 0.095 and an R² value of 0.810 demonstrating strong predictive power and robustness. SHAP based feature importance analysis revealed that pre-existing psychological conditions especially drug addiction and anxiety before COVID-19 were the most influential predictors of post-COVID anxiety. Other critical factors included low energy, age, sleep quality and vaccination status. This study highlights the potential of combining ensemble learning with explainable AI to provide reliable predictions and clinically relevant insights, supporting early detection and targeted mental health interventions for recovered COVID-19 patients.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Rakibul Hasan Nirob
AU  - Francis Rudra D. Cruze
AU  - Md. Faruk Hosen
AU  - Fizar Ahmed
AU  - Md. Nasimul Kader
PY  - 2026
DA  - 2026/06/08
TI  - Psychological Risk Profiling for Post-COVID-19 Anxiety Using Interpretable Ensemble Learning
BT  - Proceedings of the International Conference on Intelligent Data Analysis and Applications (IDAA 2025)
PB  - Atlantis Press
SP  - 358
EP  - 370
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-664-7_25
DO  - 10.2991/978-94-6239-664-7_25
ID  - Nirob2026
ER  -