# ML for Fast Simulations

Last updated on
Sep 10, 2023

Developed a graph-based generative adversarial network, MPGAN, which proved to be state-of-the-art at simulating particle collisions. Developed as well the faster attention-based generative adversarial particle transformer (GAPT), using set transformers, and efficient and sensitive two-sample goodness-of-fit tests for validating fast simulations. Working now on extending to conditional generation and application to detector data.

## Publications

Raghav Kansal, Anni Li, Javier Duarte, Nadezda Chernyavskaya, Maurizio Pierini, Breno Orzari, Thiago Tomei
(2023).
Evaluating generative models in high energy physics.
Phys. Rev. D.

Mary Touranakou, Nadezda Chernyavskaya, Javier Duarte, Dimitrios Gunopulos, Raghav Kansal, Breno Orzari, Maurizio Pierini, Thiago Tomei, Jean-Roch Vlimant
(2022).
Particle-based fast jet simulation at the LHC with variational autoencoders.
Machine Learning: Science and Technology.

Raghav Kansal, Javier Duarte, Hao Su, Breno Orzari, Thiago Tomei, Maurizio Pierini, Mary Touranakou, Jean-Roch Vlimant, Dimitrios Gunopulos
(2021).
Particle Cloud Generation with Message Passing Generative Adversarial Networks.
NeurIPS 2021.

Farouk Mokhtar, Raghav Kansal, Daniel Diaz, Javier Duarte, Joosep Pata, Maurizio Pierini, Jean-Roch Vlimant
(2021).
Explaining machine-learned particle-flow reconstruction.
ML and the Physical Sciences Workshop @ NeurIPS 2021.

Breno Orzari, Thiago Tomei, Maurizio Pierini, Mary Touranakou, Javier Duarte, Raghav Kansal, Jean-Roch Vlimant, Dimitrios Gunopulos
(2021).
Sparse Data Generation for Particle-Based Simulation of Hadronic Jets in the LHC.
LatinX in AI Research Workshop @ ICML 2021.

Raghav Kansal, Javier Duarte, Breno Orzari, Thiago Tomei, Maurizio Pierini, Mary Touranakou, Jean-Roch Vlimant, Dimitrios Gunopulos
(2020).
Graph Generative Adversarial Networks for Sparse Data Generation in High Energy Physics.
ML and the Physical Sciences Workshop @ NeurIPS 2020.

## Talks and Posters

MODE Workshop on Differentiable Programming for Experiment Design,
Machine learning for particle physics simulations

UCI Particle/Astro ML Seminar,
Generative transformers and how to evaluate them (+ Lorentz-equivariant networks)

PHYSTAT-2samples Workshop,
Applications of two-sample goodness-of-fit tests to deep generative models

US CMS Annual Collaboration Meeting (CMS-only),
Multivariate goodness of fit testing for evaluating HEP generative models

CMS Statistics Committee (CMS-only),
Multivariate goodness of fit testing for evaluating HEP generative models

MITP Machine Learning for Particle Physics Workshop,
Particle Cloud Generation with Message Passing GANs

JMU Artificial Intelligence and Machine Learning Seminar,
Graph GANs for High Energy Physics Data Generation

BIDS Deep Generative Models for Fundamental Physics Meeting,
Graph GANs for High Energy Physics Data Generation

NeurIPS 2020 Machine Learning and the Physical Sciences Workshop,
Graph GANs for Sparse Data Generation in High Energy Physics

Inter-Experimental LHC Machine Learning Working Group Meeting,
Sparse Data Generation with Graph GANs