Skip to content

marcocarotta/CMEPDA

 
 

Repository files navigation

CMEPDA

Documentation Status

Repository for the Computing Methods for Experimental Physics and Data Analysis course exam project.

Luca Callisti, Marco Carotta, Igor Di Tota

Introduction

The main purpose of this project is to implement a lossy compression, using flow-based generative models. The realization is obtained using Affine Autoregressive Flows, an example of Normalizing Flows. As an application of Normalizing Flows, it was also shown how new data can be generated from the Gaussian distributions into which the original data are mapped.

This work was inspired by the Baler tool development, where a lossy compression is realized through an autoencoder. For this reason the datasets used were the same, so that a comparison could be made.

The document Description_of_the_project_Callisti_Carotta_DiTota contains the more detailed explanation, along with the results obtained, and is available in this repo.

About

A lossy compression tool via flow-based models, optimized for HEP datasets.

Resources

License

Stars

Watchers

Forks

Contributors

Languages

  • Jupyter Notebook 99.9%
  • Python 0.1%