Prediction Analysis of Sleep Disorders Using Machine Learning-Based Techniques

Abstract

Sleep is crucial indicator for an individual. Poor sleep quality has serious implication for health. This condition is often triggered by high work pressure and imbalance between work and rest time. While previous research with similar topic has been conducted, it has not comprehensively elucidated the key factors influencing sleep disorders. Therefore, this study conducts more in-depth analysis of factors contributing to sleep disorders including; gender, age, occupation, sleep duration, quality of sleep, physical activity level, stress level, BMI, heart rate, and daily steps. Subsequently, we employ Machine Learning (ML) techniques to investigate further sleep disorders. The ML models include: Naïve Bayes (NB), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Logistic Regression (LR), Convolutional Neural Network (CNN), dan Long Short-Term Memory Network (LSTM). The objective is to assess the effectiveness of ML model implementation based on information from data and the significance of specific factors in predicting sleep disturbances. The results of this study indicate that the combination of the LR model with Chi-Square achieved the highest average F1 score, which was 84.75%, in sleep disorder classification. The research comprises several stages: (1) Data collection, (2) Pre-processing of the collected data, and (3) Training models capable of processing data for evaluation to understand the contribution of indicators to sleep disorder predictions. The findings of this study provide insights into the effectiveness of the constructed models in predicting sleep disorders