🏋️‍♂️ Fitlife Health and Fitness Prediction Web App

This project analyzes health and fitness patterns using a dataset that includes physiological, activity-based, and lifestyle-related metrics. The aim is to develop a predictive Web Application that helps users estimate:

• 🔢 Weight
• 💪 Fitness Level
• 🔥 Calories Burned
• 🩺 Health Condition

---

🚀 Project Goals

• Analyze and clean health-related data.
• Apply machine learning models to make accurate predictions.
• Balance imbalanced datasets using techniques like SMOTE and resampling.
• Deploy a user-friendly web application for real-time health predictions.

---

🧠 Features

• Standardization and label encoding for preprocessing.
• Classification and regression models using XGBoost.
• Class balancing using SMOTE and Resampling.
• Visualization tools like Seaborn and Matplotlib.
• Performance metrics such as Accuracy, F1 Score, RMSE, and R².

---

📂 Dataset

The dataset includes features such as:

• Age, Gender
• BMI, Weight
• Sleep Time, Steps
• Heart Rate, Stress Levels
• Daily Activity Metrics
• Lifestyle Inputs (e.g., diet, hydration)

Data preprocessing includes handling missing values, encoding categorical variables, and feature scaling.

---

🛠️ Technologies Used

• Python, Pandas, NumPy
• Scikit-learn, XGBoost, imblearn
• Matplotlib, Seaborn for data visualization
• Jupyter Notebook for experimentation

---

🔍 Model Workflow

1. Load and preprocess the data
2. Balance classes using resampling and SMOTE
3. Split data into training and testing sets
4. Train models (XGBoost Classifier/Regressor)
5. Evaluate using classification/regression metrics
6. Save models and integrate with web interface

---

📊 Results

• Classification models used for predicting health condition
• Regression models for weight, calories burned and fitness level

Source

https://www.kaggle.com/datasets/jijagallery/fitlife-health-and-fitness-tracking-dataset