Jet Tagging with Machine Learning

Jet_Tagging is a repository for classifying particle jets using the JetNet dataset. It applies machine learning techniques to differentiate between jet origins, such as quarks and gluons, based on their substructure properties.

🚧 This project is under active development! Expect updates and improvements. 🚧

Key Features:

Model Architectures: Includes implementations of various neural network architectures.

Training Pipelines: Provides scripts and configurations to train models on jet datasets, with support for hyperparameter tuning and performance monitoring.

Evaluation Metrics: Offers tools to assess model performance using standard metrics in particle physics, ensuring reliable and interpretable results.

Results and Visualizations

Below are key performance visualizations from binary and multi-class classifiers.

Confusion Matrices

These confusion matrices illustrate the classification performance of the models.

Binary Classifier (Quark vs Gluon)	Multi-Class Classifier

Directories and Files

• README.md: Project documentation and instructions.
• config/: Configuration files for various setups.
• data/: Scripts and utilities for handling datasets.
• evaluation/: Evaluation scripts and results.
• images/: Generated figures and plots.
• launch_project.py: Script to launch the project.
• models/: Directory for model definitions and architectures.
• scripts/: Miscellaneous scripts for different tasks.
• summary/: Summaries and logs of training runs.
• trainers/: Training scripts and utilities.
• utils/: Utility functions and helpers.

Getting Started

Prerequisites

• Jetnet (https://github.com/jet-net/JetNet.git)
• Python 3.10
• Pytorch 2.4.1

Download

1. Clone the repository:

   git clone [email protected]:CarlosSarasty/Jet_Tagging.git

2. Navigate to the project directory:

   cd Quark_Gluon_Classifier

Usage

To launch the project, run:

python launch_project.py config/config.yaml

Citations

JetNet

• JetNet GitHub Repository: JetNet
• Paper:
  Kansal, R., Duarte, J., Su, H., Orzari, B., Tomei, T., Pierini, M., Touranakou, M., Vlimant, J-R., & Gunopulos, D.
  Particle Cloud Generation with Message Passing Generative Adversarial Networks,
  Advances in Neural Information Processing Systems, Vol. 34, 2021.
  Paper Link | arXiv:2106.11535