An Optimal Random Forest Model for occupational frostbite prediction in the grid workplace

Abstract:
Frostbite is one of the most common diseases in occupational environment. Because of its rapid onset and great harm to human body, there is an urgent need for a machine learning model that can accurately predict the risk of frostbite in workers. In this paper, we introduce a tree model to predict the risk of frostbite incidence using the variable related to working environments. We found the random forest tress outperformed the linear model, which achieved an accuracy over 86.90%. Our model offers a novel approach to for the prevention of occupational frostbite.