Galvanic skin response based stress level detection using supervised machine learning approach

Stress increases the risk of heart problems and can be a hindrance in daily chores. In this study, we have acquired values using a Galvanic Skin Response (GSR) sensor. GSR is a measurement of ongoing changes in the electrical properties (conductivity) of the skin brought on by variations in the activity of human perspiration. This theory is predicated on the idea that sweat gland activity alters the skin’s resistance. Only the resistance was returned in order to calculate the conductivity. We have analysed these values by detecting the level of stress utilizing supervised machine learning algorithms. A person’s physiological state can be utilised to determine whether or not they are stressed. They are used to forecast whether a person’s stress levels will rise or fall while they carry out activities like exercising or tensing. We have developed a model using a custom-made dataset to detect the stress levels of an individual. The model was tested using multiple classifiers and makes accurate predictions to determine the stress level given a single input value, obtained from the GSR sensor.