Analysis of obesity data using Machine Learning Techniques

Abstract

Backgrounds/Objectives:  Obesity is a chronic health condition marked by the excessive build-up of body fat, typically caused by an imbalance between the calories consumed and the energy used by the body. The World Health Organization classifies it as a major global health challenge that affects individuals of all ages. Beyond its visible effects on body shape, obesity significantly increases the likelihood of developing a range of serious illnesses. Extra fat in the body can interfere with normal metabolic functions, strain vital organs, and disturb hormonal balance. This makes obesity a key contributor to diseases such as heart disorders, type 2 diabetes, breathing problems, certain cancers, liver damage, and joint issues. Gaining awareness of these risks plays an important role in prevention, early detection, and maintaining long-term health. Obesity is a growing health issue worldwide, driven by factors such as unhealthy eating habits, physical inactivity, and lifestyle changes. Early identification of individuals at risk is essential for managing and preventing obesity-related complications.

Methods/Statistical Analysis: This research employs four classification algorithms—J48, Classification and Regression Trees (CART), Alternating Decision Tree (ADTree), and Best First Tree (BF Tree)—to analyze data related to obesity. These machine learning techniques are used to classify and predict obesity by examining various health and lifestyle variables.

Findings:The study compares the accuracy and effectiveness of the different classification methods in detecting obesity. It also identifies important factors that contribute to obesity based on the analysis of the dataset.