Pedro E. Rivera-Cardona

MIT Department: Physics
Faculty Mentor: Prof. Jesse Thaler
Undergraduate Institution: University of Puerto Rico, Mayaguez

Website: GitHub Page, Twitter
Research Poster
Lightning Talk

Biography


I am a rising senior majoring in physics with a minor in pure mathematics. I am a research student at the UPRM Compact Muon Solenoid research group. My goal is to pursue a Ph.D. in theoretical physics and work on field theory or Beyond the Standard Model physics. During graduate school, I plan to participate in diversity and inclusion groups to promote interest in physics to fellow Hispanics/LatinX. In my free time, I enjoy walking with friends, drinking coffee, reading, and sightseeing. During and after the MSRP, I work with Jesse Thaler on an extension to Energy Flow Networks, a machine learning model to analyze collider events.

 

2021 Abstract

 

Implementation of U(1) Group Symmetry on Energy Flow Networks for Full
Event Analysis

Pedro E. Rivera-Cardona1, Jesse Thaler2, Rikab Gambhir3
1Department of Physics, University of Puerto Rico, Mayaguez
2, 3Center for Theoretical Physics, Massachusetts Institute of Technology

The Large Hadron Collider (LHC) is a high-energy particle accelerator that produces multitudes of particles per collision. Machine learning techniques are used on LHC data to understand particle distributions and obtain insight into how experimental measurements relate to theoretical frameworks. Energy Flow Networks (EFNs) study the unordered, variable-length sets of particles from collision events. This architecture is used to analyze and learn from collider events and other particle physics phenomena. EFNs parametrize infrared- and collinear- safe observables by a learnable per-particle function Φ and latent space function F. Previously, EFNs have been used to study single jets at a time. We present an extension to the EFNs architecture with U(1) cylindrical symmetry. U(1) cylindrical symmetry allows for full event analysis with manifest periodicity. This was  achieved  by  implementing  a  new  initial  layer  Φ!, which avoids altering the dataset. Φ! imposes a coordinate transformation to the azimuthal angle coordinate Φ, manifesting periodicity.

Finally, the extension will enable a Fourier decomposition of events to cosines and sines. Fourier decomposition will allow us to extract information on an understandable basis. After the implementation, we deploy CMS Open Data for further analysis, such as quark/gluon discrimination and top jet tagging.