Scholarlink

In the present era, an upsurge in sleep disorder cases is a concerning issue. The classification of sleep disorders using technology has the promising potential to redefine the way the healthcare industry works in this sector as manual detection is prone to human errors. The timely identification and detection of sleep disorders will improve human health exponentially. In this paper, several machine learning techniques were executed whose performances were studied and compared in their respective ability to segregate sleep disorders. The openly accessible dataset on Kaggle, namely ‘Sleep, Health and Lifestyle dataset’ has been used in this research. The outlier removal has also been done along with label and categorical encoding while preprocessing the data. The algorithms namely K-nearest Neighbors, Naive Bayes, Support Vector Machine, Gradient Boosting Classifier, Decision Tree, and Random Forest, have been studied. The corresponding accuracies of these algorithms are 91.67%, 91.67%, 87.50%, 94.44%, 93.06%, and 94.44%. Other performance metrics have also been discussed in the paper. Furthermore, the feature importance of different features in the dataset has been examined and BMI (Body Mass Index) is identified as the most significant feature while predicting the results. The maximum accuracy is achieved by Random Forest and Gradient Boosting Classifier model in this paper.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Empirical Comparison of Sleep Disorder Prediction Using Machine Learning Techniques

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Akshita Jha,
  • Ritu Bhardwaj,
  • Payal Jha,
  • Shweta Jindal

摘要

In the present era, an upsurge in sleep disorder cases is a concerning issue. The classification of sleep disorders using technology has the promising potential to redefine the way the healthcare industry works in this sector as manual detection is prone to human errors. The timely identification and detection of sleep disorders will improve human health exponentially. In this paper, several machine learning techniques were executed whose performances were studied and compared in their respective ability to segregate sleep disorders. The openly accessible dataset on Kaggle, namely ‘Sleep, Health and Lifestyle dataset’ has been used in this research. The outlier removal has also been done along with label and categorical encoding while preprocessing the data. The algorithms namely K-nearest Neighbors, Naive Bayes, Support Vector Machine, Gradient Boosting Classifier, Decision Tree, and Random Forest, have been studied. The corresponding accuracies of these algorithms are 91.67%, 91.67%, 87.50%, 94.44%, 93.06%, and 94.44%. Other performance metrics have also been discussed in the paper. Furthermore, the feature importance of different features in the dataset has been examined and BMI (Body Mass Index) is identified as the most significant feature while predicting the results. The maximum accuracy is achieved by Random Forest and Gradient Boosting Classifier model in this paper.