AUTHOR COPY Journal of Intelligent & Fuzzy Systems 44 (2023) 1085–1097 DOI:10.3233/JIFS-221588 IOS Press 1085 Effects of long-term exercise training on physiological signals and personality traits in women in law enforcement Remya George a,∗ , Reshma Jose a , K. Meenakshy b , T. Jarin c and S. Senthil Kumar d a Department of Biomedical Engineering, Sahrdaya College of Engineering and Technology, Thrissur, Kerala, India b Department of Electrical and Electronics Engineering, Government Engineering College, Thrissur, Kerala, India c Department of Electrical and Electronics Engineering, Jyothi Engineering College, Thrissur, Kerala, India d Department of Electrical and Electronics Engineering, New Prince Shri Bhavani College of Engineering and Technology, Chennai, Tamil Nadu, India Abstract. Law enforcement teams across the globe experience the highest occupational stress and stress-related diseases. Physical exercise and an active lifestyle are recommended as part of their profession to equip them to fight stress and related health adversities. The research is carried out using objective measures of Heart Rate Variability (HRV), Electro Dermal Activity (EDA), Heart Rate Recovery (HRR), and subjective questionnaires. HRV was generated with an electrocardiogram (ECG) signal acquired using NI myRIO 1900 interfaced with the Vernier EKG sensor. HRR was acquired with the help of a Polar chest strap exercise heart rate monitor and EDA acquisition was carried out with Mindfield E-Sense electrodes. Then statistical features are extracted from the collected data, and feed to the AQCNN (Aquila convolution neural network) classifier to predict the stress. Signal analyses were done in Kubios 4.0, Ledalab V3.x in a MATLAB environment. The results pointed out that exercise training is effective in increasing the vagal tone of the Autonomic Nervous System (ANS) and hence improves the recovery potential of the cardiovascular system from stress. The proposed AQCNN method improves the accuracy by 95.12% which is better than 93.13%, 85.36% and 80.13% from Statistical technique, CNN and ML-SVM respectively. The findings have the potential to influence decision-making in the selection and training of recruits in high-stress positions, hence optimizing the cost and time of training by identifying maladaptive recruits early. Keywords: Exercise training, ANS adaptation, machine learning, stress-recovery, heart rate variability, heart rate recovery, electrodermal activity 1. Introduction In the current worldwide environment, police forces are thought to have the greatest levels of occupational stress and stress-related disorders [1]. According to the hypothesis of Cross-stressor adapta- ∗ Corresponding author. Remya George, Department of Biomedical Engineering, Sahrdaya College of Engineering and Technology, Thrissur, Kerala 680684, India. E-mail: remyageorge92@gmail.com. tion, regular exercise can affect physiological systems and help to soften the response to both physi- cal and psychological stresses [2, 3]. Furthermore, scientific literature underlines that increased phys- ical activity or high fitness is likely to improve stress tolerance, which benefits leading a healthy lifestyle [4]. Consistent, cyclic delivery of varied loads of training is regarded as the key to physio- logical adaptation in police training camps all over the world. However, due to imbalances in genetic ISSN 1064-1246/$35.00 © 2023 – IOS Press. All rights reserved.