Obesity Prediction Using Machine Learning

Project Overview

This project focuses on predicting obesity levels based on a variety of physiological, behavioral, and lifestyle factors using supervised machine learning. The classification model categorizes individuals into different obesity levels using structured health-related data, aiming to support proactive healthcare decisions and awareness.

---

Technical Highlights

• Dataset:

  • 2,111 samples and 17 features
  • Includes both numerical (age, BMI, water intake, physical activity) and categorical (gender, diet habits, alcohol, transport) attributes
  • Target label: Obesity (multiclass categories)

• Exploratory Data Analysis:

  • Analyzed distributions of variables like age, gender, water/alcohol intake
  • Visualized correlations between numerical features (e.g., BMI vs Obesity)
  • Investigated lifestyle factors (e.g., transportation, smoking, eating habits)

• Preprocessing:

  • Dropped duplicates
  • Feature engineering (e.g., calculated BMI)
  • Encoding: Ordinal, One-Hot, and Label Encoding
  • Feature scaling using Standard Scaler

• Model Building:

  • Used GradientBoostingClassifier for multiclass classification
  • Achieved training accuracy of 100% and testing accuracy of 97%

• Evaluation:

  • Used confusion matrices and performance metrics (Accuracy, Precision, Recall, F1-score)
  • Achieved strong generalization performance across all obesity classes

---

Key Use Cases

• Predictive tool for obesity level classification in health-tech applications
• Enables personalized fitness or diet recommendations
• Supports health insurance assessments based on lifestyle risk factors
• Useful in preventive healthcare analytics

---

Installation

Clone the repository

git clone https://github.com/BhaveshBhakta/Obesity-Prediction-Using-ML.git
cd Obesity-Prediction-Using-ML

---

Contributions

Contributions are welcome! If you’d like to improve model performance, add new visualizations, or integrate the project with a web interface.