Exploring Sleep Paralysis Phenomenon Through Machine Learning: An Analytical Study

Abstract

Sleep paralysis is when you're awake but powerless to move. Although the majority of occurrences are linked to extreme terror and some potentially clinically significant sufferings are connected with the case; little is known about it. The present state of research on the connection between sleep paralysis and overall sleep will be examined in this study. Several studies have linked an increased risk of sleep paralysis to low-quality sleep.. This could occur in between sleeping and waking. The issue is tackled in the context of sleep paralysis prediction using machine learning. There are multiple attributes in the dataset (label Data). Every exam has a unique set of characteristics and expected results. We carried out an accuracy test of posture recognition by the system to evaluate its validity. Predicting sleep paralysis has been done using machine learning techniques. A comparison of sleeping positions with the findings indicated a correlation between paralysis and poor quality of sleep. Poor sleep quality is correlated with sleep paralysis. In the provided dataset, the Random Forest model predicts sleep paralysis with the highest accuracy 92.3%, K-nearest neighbors' classifier has an accuracy rate of 91.0% 91.9%, and 90.4% using Decision Tree and naïve Bayes.