cern_ml's Introduction

This project was done in the framework of a semester project at CERN in the first year of the Master. More details on the background and results are available in this paper.

Structure

├── create_sample.py        # Script used to generate the model on SCITAS supercomputer  
├── train_model.py          # Script used to train the model on SCITAS supercomputer.
├── sample_methods.py       # All the methods related to the creation of the training dataset 
├── NN_methods.py           # All the methods related to the training of the GNN models
├── results_analysis.ipynb  # Analysis of predictions made by GNN models
├── Docs                    
│   ├── requirement.txt     # Packages required to run the project
│   └── paper.pdf           # Report of this project
│ 
└── Data
    ├── Analysis_of_results # Folder with all the dataframe generated during the analysis part in .pkl format
    └── CERN_DAs            # Folder with all the DAs data in .npz format

Installation

To install and run the project, you first need to download all the libraries from requirements.txt. Then you need to download the real DAs dataset to this link and put it in the folder Data/. The training and test sets are too heavy to be uploaded on a drive. They have to be generated from scratch with the provided script create_sample.py.

Recommend Projects