Analysis of Machine Learning Models for Gamma-Ray Detection

Description

This independent research project leverages Machine Learning techniques in Python to classify cosmic gamma-rays and hadronic cosmic-ray backgrounds in telescopic data. The study utilizes various algorithms such as Support Vector Machines, Logistic Regression, Naive Bayes, K-Nearest Neighbors, and Neural Networks for the classification task. The project involves feature engineering, regularization techniques, and emphasizes data pre-processing, model optimization, and results interpretation.

Table of Contents

• Abstract
• Introduction
• Setting up the Dataset
• Training, Validation, and Testing Datasets
• Classification Report Explanation
• KNN
• Naïve Bayes
• SVM
• Logistic Regression
• Logistic Regression vs. Neural Network
• Neural Network
• EDITED Neural Network
• Discussion
• Conclusion
• Technology Stack
• Code Snippet

Installation

1. Clone the GitHub repository.
2. Ensure you have Jupyter Notebook installed.
3. Navigate to the notebook directory and run jupyter notebook to open the notebook.

Usage

Open the notebook in Jupyter and execute the cells in sequence. Adjust model parameters as needed for further experimentation.

Technology Stack

• Core Libraries: numpy, pandas, matplotlib
• Machine Learning Libraries: sklearn.preprocessing, imblearn.over_sampling, sklearn.neighbors, sklearn.metrics, sklearn.naive_bayes, sklearn.svm, sklearn.linear_model

Code Snippet

import numpy as np
import pandas as pd
from sklearn.preprocessing import StandardScaler

# Sample code to standardize data
data = pd.read_csv('data.csv')
scaler = StandardScaler()
standardized_data = scaler.fit_transform(data)

Contributors

• Syed Bilal Afzal: Primary researcher and author.
• Jicai Pan: Project supervisor.

References

• Stanford University Notes on Classification
• Support Vector Machines (SVM) Notes
• YouTube Machine Learning Playlist

Acknowledgments

Special thanks to online course materials, textbooks, Stanford University Notes, YouTube, Codecademy, public GitHub pages, and Classification ML textbooks for providing valuable resources that supported this research.

---

Note: The code snippet provided is a simplified example for the sake of illustration. Please refer to the notebook for detailed implementations.