LSTM Based Mental Stress Level Detection using Wearable Sensor Devices

Nowadays, mental stress is a severe problem, particularly among teenagers. The age group that was previously thought to be the most carefree is now under much strain. Nowadays, increased stress causes many new issues, including depression, suicide, heart attack, and stroke. Many physical health issues related to stress can be avoided if mental stress is detected beforehand. When a person is stressed, there are noticeable changes in several bio-signals such as thermal, electrical, impedance, acoustic, optical, and others, and stress levels may be determined using these bio-signals. The paper proposes various machine learning and deep learning algorithms for stress detection on persons using wearable sensing devices, which can help people avoid various stress-related health problems. The data are taken from the WESAD dataset. The paper uses classification algorithms, SVM, DT, kNN and deep learning models; Long Short-Term Memory (LSTM) is applied, and confusion metrics parameters are used to measure performance in terms of accuracy, precision, recall, f1-score