ASSESSING THE EFFECT OF LONG TERM PHYSICAL TRAINING AND CLASSIFICATION OF TRAINING STATUS USING HRV AND HRR OF FEMALE POLICE RECRUITS

摘要

The changes in Heart Rate Variability (HRV) and Heart Rate Recovery (HRR) coincide well with the changes in physicaltraining status in patient populations. But a deep probing in this area reveals that enough attention has not been paid so far to thehealthy population, especially those who undergo deliberate training. This study was conducted to quantify the effect of ninemonths of basic police training in HRV/HRR among a sample of women police recruits of Kerala state. Consequently, the trainingwas found to be effective in altering the parasympathetic and nonlinear control of the cardiovascular system. The statisticalanalysis using dependent sample t-test showed that there was a significant alteration of linear and nonlinear HRV measures andHRR features prior to post-training. The study also investigated the discriminatory potential of five-minutes each of supine HRVand post-exercise HRR in classifying the recruits into trained and untrained status. The optimal HRV/HRR feature set that coulddiscriminate the training status with the highest accuracy were identified by using Genetic Algorithm-Artificial Neural Network(GA-ANN) and Genetic Algorithm-Support Vector Machine (GA-SVM) based wrapper functions. A reduction in HRV feature setto 50% and HRR feature set to 68.5% was found using the GA optimization. While classifying, the SVM classifier outperformedthe ANN with maximum accuracy (89.7%) using the reduced feature set of HRR. The results promise objective selection of welltrainedcandidates to professions which demand high physical fitness.