EPE Seminar: Aishik Ghosh

Tuesday Nov 12, 2024, 4:45 PM → 5:45 PM US/Pacific

PAT C-421

Henry J Lubatti (lubatti@uw.edu), Henry Lubatti (University of Washington (US)), Quentin Buat (University of Washington (US)), Shih-Chieh Hsu (University of Washington Seattle (US))

Title: High-dimensional statistical inference with AI in Higgs boson physics and beyond

Abstract: Neural Simulation-Based Inference (NSBI) is a powerful class of machine learning (ML)-based methods for statistical inference that naturally handle high dimensional data, avoiding the need to design low-dimensional summary statistics (eg. histograms). I show how such methods are promising for measurements at the Large Hadron Collider (LHC), where low-dimensional representations of the data fail to capture the full complexity of the quantum interference effects at play, and this leads to a loss of precision in the measurement. We then develop a framework in the ATLAS experiment that enables the application of NSBI to a full-scale Higgs width analysis, leading to the most precise measurement of the Higgs width by the experiment to date. This work involves the development of several innovative solutions to introduce uncertainty quantification and enhance robustness and interpretability in neural inference. The developed method It is a generalisation of the standard frequentist statistical inference framework used in particle physics, and therefore will benefit a wide variety of applications in experiments across the field. I will also show an example of how I applied the same concepts for high-dimensional Bayesian inference in astrophysics.


Bio:  Dr. Aishik Ghosh is a particle physicist at UC Irvine and Berkeley Lab where he studies Higgs physics at the ATLAS experiment using novel statistical and  uncertainty quantification methods, that he develops. His current efforts focus on the Higgs width measurement, and he is also working on calorimeter simulation and trigger algorithms. He obtained his PhD from the Université Paris-Saclay for developing simulation-based-inference methods for the Higgs width measurement and deploying the first deep generative model for fast simulation of the ATLAS detector. Dr. Ghosh also works with the Organisation for Economic Co-operation and Development on matters of AI and science policy.

This event is co-hosted by A3D3.

Zoom: https://cern.zoom.us/j/66336943729?pwd=hhXnNFyjDbPnOj5Hu6UVf3c50qIbru.1
pass code: 69746646

4:45 PM → 5:25 PM Higgs physics with Neural Simulation-Based Inference at the LHC and in ALTAS

Speaker: Aishik Ghosh (University of California Irvine (US))

Recording

UW_12Nov2024.pdf