Relationship between Firefighter Physical Fitness and Special Ability Performance: Predictive Research Based on Machine Learning Algorithms

Abstract

Firefighters require a high level of physical fitness to meet the demands of their job. The correlations and contributions of individual physical health parameters to the tasks of firefighting would enable firefighters to focus on the effects of specific physical conditions during their physical training programs. Therefore, the purpose of the present study was to identify the relationships between various physical health parameters (weight, maximum oxygen uptake, body fat percentage, upper body muscular power and lower body muscular power) and performance on simulated firefighting ability tasks, which included a set of seven tasks (rope climb, run 200 m round trip with load, 60 m carrying a ladder, climb stairs with load, evacuation of 400 m with supplies, run 5 km with an air respirator, run 100 m with the water hose). Through use of a partial least-squares regression (PLSR) algorithm to analyze the linear correlation, we revealed the change in various training performances of specific ability tests with physical fitness parameters. The present study demonstrated significant relationships among physical health parameters and performance on simulated firefighting ability tasks, which also represent that those parameters contributed significantly to the model’s predictive power and were suitable predictors of the simulated firefighting tasks score.