Gamma-Ray Classification Using Machine Learning

Project Overview

This project focuses on classifying high-energy gamma particles using machine learning techniques. The dataset used is the MAGIC Gamma Telescope dataset, which contains measurements recorded by imaging atmospheric Cherenkov telescopes. The goal is to build a predictive model capable of distinguishing gamma signals from hadronic background noise.

The project covers machine learning workflow including:

• Data loading and preprocessing
• Exploratory Data Analysis (EDA)
• Feature engineering
• Model training and evaluation
• Hyperparameter tuning
• Dashboard visualization

---

Objectives

• Clean and preprocess the MAGIC Gamma Telescope dataset
• Explore relationships between telescope measurement features
• Train classification models for gamma-ray prediction
• Evaluate model performance using multiple metrics
• Save the best-performing model
• Build an interactive dashboard for visualization

---

Dataset

Dataset: MAGIC Gamma Telescope Dataset

Features include:

• fLength
• fWidth
• fSize
• fConc
• fConc1
• fAsym
• fM3Long
• fM3Trans
• fAlpha
• fDist

Target Variable:

• g → Gamma signal
• h → Hadron background noise

---

Technologies Used

• Python
• Pandas
• NumPy
• Matplotlib
• Seaborn
• Scikit-learn
• Joblib
• Tableau
• Git & GitHub

---

Data Preprocessing

The preprocessing stage included:

• Assigning column names
• Handling missing values
• Encoding target labels
• Feature scaling
• Train-test splitting

---

Machine Learning Models

The following models were explored:

• Logistic Regression
• Random Forest Classifier
• Decision Tree Classifier
• Support Vector Machine (SVM)

Hyperparameter tuning was performed to improve performance.

---

Evaluation Metrics

Models were evaluated using:

• Accuracy Score
• Precision
• Recall
• F1 Score
• ROC Curve
• Confusion Matrix

Example visualization outputs include:

• Feature importance plots
• ROC curves
• Correlation heatmaps
• Class distribution charts

---

Model Saving

The best-performing model was saved using Joblib:

import joblib

joblib.dump(model, "best_model.pkl")

---

Running the Project

Clone Repository

git clone <your-repository-link>
cd gamma-ray-classification

Install Requirements

pip install -r requirements.txt

Run Dashboard

link.md

---

Dashboard Features

The dashboard includes:

• Dataset overview
• Feature visualizations
• Prediction interface
• Model performance metrics
• Interactive charts

---

Results

The final trained model achieved strong classification performance in distinguishing gamma particles from hadronic events. The project demonstrates how machine learning can support astrophysics research and high-energy particle analysis.

---

Future Improvements

• Deep learning implementation
• Real-time prediction API
• Cloud deployment

---