Injury prediction for Canadian mineral exploration using machine learning and time series

Abstract

The mineral exploration industry, vital to Canada's economy, is also one of the most hazardous sectors, with workers facing significant risks of severe injuries or fatalities. This research focuses on improving occupational health and safety by predicting nature of injury within the industry. Utilizing an extensive dataset from the Prospectors and Developers Association of Canada (PDAC), the study applies advanced machine learning (ML) techniques to forecast injury severity across four classes. Eight ML methods, including Support Vector Machine, Convolutional Neural Network, Bayesian Neural Network, and more, along with four time series methods, were used. The findings offer critical insights into injury patterns, enabling the development of enhanced safety protocols. By providing quantitative predictions, this research aims to reduce injury rates and improve safety measures in the mineral exploration sector, contributing to a safer working environment.