Stress and Anxiety Prediction Using Machine Learning: A Benchmark Study of Key Algorithms

Abstract

This research examines how effective machine learning can be in forecasting stress and anxiety levels, with an emphasis on evaluating commonly used algorithms. With stress and anxiety impacting mental well-being on a large scale, predictive modeling offers a proactive approach to mental health assessment. Using physiological and behavioral data from a validated dataset, we evaluate and compare models such as Random Forest Classifier, SVM, Gradient Boost Classifier and Logistic Regression. Results indicate distinct performance variances across models, providing insights into algorithm suitability for real-world applications in mental health prediction, paving the way for personalized intervention strategies.