MAGIC Gamma Telescope Classification - Assignment 2

This project contains a simple, step-by-step notebook implementation for the assignment requirements using the MAGIC dataset.

Files

• src\assignment2_classification.ipynb - main notebook
• data\raw\magic04.data - raw dataset input
• data\processed\processed_balanced.csv - balanced dataset
• data\processed\train.csv - training split (70%)
• data\processed\test.csv - testing split (30%)

Requirements

• Python 3.10 or newer
• pandas
• matplotlib
• scikit-learn
• jupyter

Install:

pip install pandas matplotlib scikit-learn jupyter

Run

1. Put the raw file in data\raw\magic04.data.
2. Open src\assignment2_classification.ipynb.
3. Run all cells in order.

Notebook workflow

1. Load the raw dataset with explicit column names.
2. Balance classes by downsampling g to match h.
3. Split data into stratified train/test (70/30).
4. Tune n_estimators for AdaBoost and Random Forest using cross-validation.
5. Train and evaluate:
   • Decision Tree
   • Naive Bayes
   • AdaBoost
   • Random Forest
6. Report accuracy, precision, recall, F1-score, and confusion matrix for each model.

Output

After running the notebook, processed CSV files are saved to data\processed\.