Predicting fatal fall from heights accidents using random forest classification machine-learning model
My primary flagship publication. Seven models were tested using 3,321 accident scenarios. Random Forest was selected for its predictive performance and ability to reveal the contribution of site, management, individual and agent factors.
- 3,321 structured accident scenarios
- Seven classification models evaluated
- Explainable factor-level prediction
- 146 Google Scholar citations at the 2026 redesign