A random forest model for exertional heat illness prediction in the power grid work place

Power grid worker may suffer Exertional heat illness (EHI) while working outdoors. There are many factors that can affect EHI, including meteorological factor, individual factor, physical work factor and other potential factors that can affect EHI. Traditional prediction methods only consider single variable and lack of in-depth mining for complex variables. Machine learning algorithms can find the key factors affecting EHI and calculate the optimal prediction model through multiple iterations within the algorithm, which is increasingly used in the prediction of various occupational diseases. In this study, a variety of machine learning models including random forest (RF), support vector machine (SVM), logistic regression (LR) and xgboost (XGC) were used to predict the risk of EHI of power grid workers. The accuracy of cross-validation was 74.5%, 68.4%, 71.1%, 69.9%, and RF had the best prediction effect.