# Lifestyle and psychosocial predictors of health resilience among Indonesian disaster preparedness cadres (TAGANA): a machine learning approach

## TL;DR

This study uses machine learning to identify factors affecting health resilience in Indonesian disaster volunteers, highlighting lifestyle and psychosocial elements.

## Contribution

The study introduces machine learning methods to analyze health resilience determinants in disaster volunteers, emphasizing novel predictive insights.

## Key findings

- Life risk is the strongest negative predictor of health resilience.
- Hope and healthy acts are significant protective factors.
- Random Forest Regression outperformed Decision Tree Regression in model stability.

## Abstract

This study investigates the determinants of health resilience among TAGANA volunteers in Indonesia using Decision Tree Regression (DTR) and Random Forest Regression (RFR). Data from 200 respondents show that life risk is the most dominant negative factor, while hope healthy acts as a strong protective factor; other variables such as disease history, dynamic experience, and physical-mental balance contribute moderately. RFR outperformed DTR in stability, confirming the robustness of the findings. The results emphasize the need for risk mitigation, psychosocial support, and lifestyle-based interventions, while international best practices provide models for strengthening volunteer resilience. This study contributes evidence-based insights for policy and training frameworks to improve TAGANA’s long-term sustainability in disaster response.