🏋️‍♂️ Active Session Predictor (Machine Learning Project)

A complete end-to-end Machine Learning project that predicts whether a gym session is Active or Not Active based on user fitness metrics. This project demonstrates core ML concepts including model building, evaluation, visualization, scaling, comparison, and hyperparameter tuning.

This project uses real-world inspired fitness data to classify whether a workout session is active (high calorie burn) or not. It is built using multiple ML techniques and provides insights through evaluation metrics and visualizations.

To predict whether a workout session is Active based on:

• Age
• Weight
• Height
• Session Duration
• Average BPM (Heart Rate)

👉 A session is considered Active if:

Calories_Burned > 400

• Logistic Regression
• Model Evaluation Metrics
• Confusion Matrix & ROC Curve
• Learning Curves (Overfitting Analysis)
• Feature Scaling (StandardScaler, MinMaxScaler)
• Model Comparison (KNN, Decision Tree, Random Forest)
• Cross Validation
• Hyperparameter Tuning (GridSearchCV)

Fitness_Model/
│
├── fitness.py # Basic Logistic Regression model
├── evaluation.py # Evaluation metrics + Confusion Matrix + ROC Curve
├── overfitting.py # Learning Curve (Overfitting/Underfitting analysis)
├── scaling.py # Feature Scaling comparison
├── multiple_models.py # Model comparison (Multiple algorithms)
├── tuning.py # Hyperparameter tuning (GridSearchCV)
│
├── gym_members_exercise_tracking_synthetic_data.csv # Dataset
│
├── output_curves/ # Saved plots
│ ├── confusion_matrix.png
│ ├── roc_curve.png
│ └── learning_curve.png
│
├── README.md
├── requirements.txt
└── .gitignore

1. Clone the repository:

git clone https://github.com/aeindri-tech/Fitness_Model.git

2. Navigate to the project folder:

cd Fitness_Model

3. Install dependencies:

pip install -r requirements.txt

Run any task file:

python fitness.py

You will be prompted to enter:

• Age
• Weight
• Height
• Session Duration
• Average BPM

👉 The model will predict:

Active Session / Not an Active Session

Confusion Matrix

ROC Curve

Learning Curve

• Achieved high accuracy (~98%)

• Evaluated using:

  • Accuracy
  • Precision
  • Recall
  • F1 Score
  • ROC-AUC

• Importance of data preprocessing
• Effect of feature scaling on model performance
• Comparing multiple models improves decision making
• Cross-validation ensures model reliability
• Hyperparameter tuning boosts performance

• Deploy as a web app (Streamlit / Flask)
• Add more features (Sleep, Diet, Steps, etc.)
• Use advanced models (XGBoost, Neural Networks)
• Improve dataset quality

This project is licensed under the MIT License.

Aeindri BTech CSE (AI) Student

If you liked this project:

• ⭐ Star the repository
• 🍴 Fork it
• 📢 Share it

This project showcases a complete beginner-to-intermediate level Machine Learning workflow and is a strong step towards building real-world ML systems.