Detecting Stress With Sleep-Related Factors via Machine Learning
by Raehyun Kim
Category: Social Science
Abstract – Stress is prevalent in modern society, and early detection is crucial. This article explores eight biological methods for detecting stress and their correlation. After data preprocessing, the LGBM classification achieved 100% accuracy, with snoring rate, temperature, and sleep duration being the top three indicators. Despite the snoring rate's potential inaccuracy due to rhinitis, it remains the most common outcome. Researchers should identify more physical and physiological symptoms to complement the biological indicators for stress detection.