PreDisM: Pre-Disaster Modelling With CNN Ensembles for At-Risk Communities

TL;DR

PreDisM is a novel ensemble deep learning approach that predicts building damages from natural hazards before they occur, aiding proactive disaster mitigation efforts.

Contribution

It introduces an ensemble of ResNets and decision trees for pre-disaster damage prediction, addressing a gap in proactive hazard damage modeling.

Findings

Model accurately estimates building vulnerability to various hazards.

Ensemble approach improves prediction robustness.

Responsive to different disaster types.

Abstract

The machine learning community has recently had increased interest in the climate and disaster damage domain due to a marked increased occurrences of natural hazards (e.g., hurricanes, forest fires, floods, earthquakes). However, not enough attention has been devoted to mitigating probable destruction from impending natural hazards. We explore this crucial space by predicting building-level damages on a before-the-fact basis that would allow state actors and non-governmental organizations to be best equipped with resource distribution to minimize or preempt losses. We introduce PreDisM that employs an ensemble of ResNets and fully connected layers over decision trees to capture image-level and meta-level information to accurately estimate weakness of man-made structures to disaster-occurrences. Our model performs well and is responsive to tuning across types of disasters and highlights the space of preemptive hazard damage modelling.