A Comparison of Multiple Machine Learning Algorithms to Predict Whole-Body Vibration Exposure of Dumper Operators in Iron Ore Mines in India

Background: This study deals with some factors that influence the exposure of whole-body vibration (WBV) dumper operators in surface mines. The also highlights approach to improve multivariate linear analysis outcomes when collinearity exists between certain factor pairs. Material and Methods: A total number 130 readings was taken from two adjacent iron ore frequency-weighted RMS acceleration used for WBV assessment operators. considered this are age, weight, seat backrest height, awkward posture, machine load tonnage, speed haul road condition. Four learning models were explored through empirical training-testing approach. Results: bootstrap regression model found be best based on performance predictability compared multiple regression, LASSO decision tree. Results revealed exposure. significant are: weight (regression coefficient ?=-0.005, p<0.001), posture (?=0.033, tonnage (?=-0.026, p<0.05), (?=0.008, p<0.001) poor condition (?=0.015, p<0.001). Conclusion: produced efficient results dataset which characterized by collinearity. is multifactorial. Regular monitoring corrective actions appropriate prevention programs including ergonomic design would increase health safety