Stress Prediction Using ML Algorithm

Abstract

In the current scenario, mental stress has become a major issue in society, impacting mental and physical well- being. Early detection of the level of stress is critical in order to intervene and avoid severe mental conditions. In this paper, a novel method of predicting stress levels using the responses obtained in the DASS-21 questionnaire is presented using the machine learning approach. In the proposed system, a Random Forest classifier is used to classify individuals based on different levels of stress severity. Data preprocessing techniques are also used to enhance the accuracy of the model, including label encoding, feature scaling, and handling missing values. In the experiment, it has been shown that the Random Forest model is capable of predicting the levels of stress and the most dominant features in the questionnaire.