IRIS-HEP Fellow: Shravan Chaudhari
Fellowship dates: May – Aug, 2021
Home Institution: Birla Institute of Technology and Science Pilani
Project: Accelerating End-to-End Deep Learning Reconstruction using Graph Neural Networks.
Currently, most of the machine learning based particle identification techniques developed by the CMS and ATLAS experiments rely on the inputs provided by the Particle Flow (PF) algorithms to convert detector level information to physics objects. Despite the very high reconstruction efficiency of PF algorithms, some physics objects fail to be reconstructed, reconstruct imperfectly or they exist as fakes. The end-to-end deep learning technique combines deep learning algorithms and low level detector representation of collision events. This project aims to implement graph neural network (GNN) based deep learning approaches to perform end-to-end tau identification. Furthermore, the developed GNN algorithm will be integrated with the existing CMS Software (CMSSW) based end-to-end deep learning framework (E2EFW).More information: My project proposal
Mentors:
-
Dr. Sergei Gleyzer (The University of Alabama)
-
Dr. Davide DiCroce (The University of Alabama)
- 6 Oct 2021 - "Graph Generative Models for Fast Detector Simulations in Particle Physics", Shravan Chaudhari, Recording: Graph Generative Models for Fast Detector Simulations in Particle Physics
Current Status
January 2023 - Research Assistant at NYU Center for Data Science
Contact me: