Tech Stack & Pipeline
- Inputs: Gender, Age, Height, Weight, Duration, Heart Rate, Body Temp.
- Algorithms: Linear Regression, Random Forest & XGBoost—selected XGBoost (R²=0.89, RMSE=25 kcal).
- Feature Engineering: BMI, HR×Duration interaction, temp-normalization.
- Validation: Stratified k-fold, hyperparameter tuning, SHAP explanations.
- Deployment (MLOps): Docker + GitHub Actions CI/CD → Render.com hosted.
App Screenshot
🔥 In fitness, every calorie counts—and so does every data point. By capturing physiological metrics (age, HR, temp) and layering advanced regression models, the CalorieBurn Predictor delivers real-time, precise burn estimates with sub-30 kcal RMSE.
I engineered the full MLOps lifecycle: data ingestion → feature pipeline → model training & explainability → containerization → automated deployment. Now users get evidence-based insights to optimize workouts and health goals.
I engineered the full MLOps lifecycle: data ingestion → feature pipeline → model training & explainability → containerization → automated deployment. Now users get evidence-based insights to optimize workouts and health goals.