🔥 Calories Burned Prediction Using ML & ANN

📌 Project Overview

This project predicts the number of calories burned during a workout based on user inputs like age, weight, height, duration, and gender. It compares traditional Machine Learning algorithms with Artificial Neural Networks (ANN) — with ANN showing superior performance.

---

---

📊 Dataset Overview

The dataset used for this project contains biometric and workout-related features of individuals. Below is a preview of the first few rows:

User_ID	Gender	Age	Height	Weight	Duration	Heart_Rate	Body_Temp	Calories
14733363	male	68	190.0	94.0	29.0	105.0	40.8	231.0
14861698	female	20	166.0	60.0	14.0	94.0	40.3	66.0
11179863	male	69	179.0	79.0	5.0	88.0	38.7	26.0
16180408	female	34	179.0	71.0	13.0	100.0	40.5	71.0
17771927	female	27	154.0	58.0	10.0	81.0	39.8	35.0

Each row represents one workout session with corresponding physiological data and calories burned.

---

📈 EDA Questions for Fitness Dataset

1. General Information

• What is the total number of unique users in the dataset?
• What is the gender distribution of users?
• Are there any missing values in any of the columns?

2. Statistical Summary

• What is the mean, median, and standard deviation of Age, Height, Weight, Duration, Heart Rate, Body Temp, and Calories?
• Are there any outliers in the numerical columns?

3. Relationship-Based Analysis

• How does workout Duration affect Calories burned?
• Is there a correlation between Heart Rate and Calories burned?
• Does Body Temperature change significantly with Age or workout Duration?

4. Grouped Comparisons

• What is the average Calories burned by Gender?
• How does the average Heart Rate vary across different Age groups?
• Is there a trend in Calories burned with increasing Age?

---

🔍 Feature Selection

What is Correlation?

Correlation is a statistical measure that expresses the extent to which two variables are linearly related. In machine learning, we use correlation to identify relationships between features and the target variable. A high correlation (close to +1 or -1) suggests a strong relationship, while a correlation near 0 indicates little to no linear relationship.

📌 Correlation Heatmap

This heatmap shows the correlation coefficients between all numerical features in the dataset.

✅ Feature Selection Insights from Heatmap

• We dropped User_ID as it is just an identifier and not useful for prediction.
• Height and Weight have a very high correlation (0.96), which indicates multicollinearity.
• To address this, we dropped Height because it has a low correlation (0.02) with the target variable Calories.

---

📈 Model Results & Comparison

Below are the evaluation metrics from various Machine Learning models and the Artificial Neural Network (ANN) used to predict calories burned. The metrics used include:

• MAE (Mean Absolute Error)
• MSE (Mean Squared Error)
• RMSE (Root Mean Squared Error)
• R² Score (Coefficient of Determination)

🔍 Linear Regression

• MAE : 8.44
• MSE : 126.47
• RMSE: 11.25
• R² Score: 96.63%

---

🔍 Support Vector Regression (SVR)

• MAE : 1.99
• MSE : 22.77
• RMSE: 4.77
• R² Score: 99.39%

---

🔍 Decision Tree Regressor

• MAE : 3.29
• MSE : 26.64
• RMSE: 5.16
• R² Score: 99.29%

---

🔍 Random Forest Regressor

• MAE : 1.72
• MSE : 7.38
• RMSE: 2.72
• R² Score: 99.80%

---

🤖 Artificial Neural Network (ANN)

• R² Score: 100.00%

The ANN model outperformed all traditional ML models and achieved perfect accuracy on the test set.

---

🧠 Real-Time Calorie Prediction

While we didn’t build a front-end interface, we implemented an interactive real-time prediction system within the Jupyter notebook. Users are prompted to input their:

• Gender (Male or Female)
• Age
• Weight (in kg)
• Workout Duration (in minutes)
• Heart Rate
• Body Temperature (in °C)

These inputs are:

• Label encoded (for Gender)
• Scaled using the same preprocessing used during training
• Passed to the trained ANN model (best_model.keras) to predict calories burned

💡 This approach provides instant, personalized calorie burn predictions using the model's learned patterns—all from within a notebook cell!

---

🧪 Libraries Used

• pandas
• numpy
• matplotlib
• seaborn
• scikit-learn
• tensorflow / keras

---

⚙️ Python Compatibility

Required Python version: 3.9 to 3.12

⚠️ TensorFlow is not compatible with versions newer than 3.12.

---

📁 Project Structure

calories-prediction/
├── data/
│   ├── exercise.csv         # Input features with User_ID
│   ├── calories.csv         # Target values (calories burned) with User_ID
│   └── (Merged during runtime using pd.merge on 'User_ID')
│
├── models/
│   ├── best_model.keras     # Trained ANN model
│   ├── gender_encoder.pkl   # Label encoder for gender feature
│   └── scaler.pkl           # Preprocessing scaler used in training
│
├── notebooks/
│   └── code.ipynb           # Full workflow: EDA → Preprocessing → Training → Prediction
│
├── requirements.txt         # Python dependencies
└── README.md                # Project documentation