Sleep Health Prediction

Keywords: Sleep Health

Background

The Sleep Health and Lifestyle dataset provides valuable information about sleep patterns, health conditions, and lifestyle factors of individuals. In this study, we investigate the application of Artificial Neural Networks (ANN) for predicting sleep disorders, specifically sleep apnea and insomnia. The dataset contains a range of features including demographic information, medical history, sleep-related measurements, and lifestyle indicators. The dataset is preprocessed to handle unnecessary columns, standardize numerical features, and encode categorical variables. We then split the dataset into training and testing sets, ensuring a balanced distribution of sleep disorder cases. The ratio in which the data was split is 75:25. As the target variable consists of three classes (none, sleep apnea, and insomnia), we employ multi-class classification techniques with ANN. After splitting the data, the target variable was converted into binary class matrix consisting of zero’s and one’s as ANN expects the output to be passed only in binary class matrix. It is done with np.utils models from Keras library. Three hidden layers were passed along with the input layer, and all of them work on ReLu activation function. The output layer performs SoftMax activation function. The optimizer used include adam, the loss applied is known as categorical cross entropy. To avoid the overfitting issue of the data, dropouts were also applied by the author. Overall, the model’s accuracy came to be 89 percent.

Aim & Objectives

Aim
To Predict the Sleep Health Using DL Techniques
Objective

Research Methodology

ANN Technique used

Software & Tools

Jupyter, Google Collab

— Preview truncated. The complete document (full chapters, references, diagrams, and appendices) is shared with clients as part of project delivery. —