School of Physics Faculty Search Colloquium Series- Dr. Aishik Ghosh

Speaker: Dr. Aishik Ghosh

Title: Probing High-Dimensional Spaces: From Theory Design to Parameter Inference in Particle and Astrophysics

Abstract: When confronted with extremely high-dimensional problems, physicists traditionally reduce the challenge to a lower dimensional representation where they can build intuition. For example, astrophysicists might summarize raw telescope images of a neutron star into regressed values for mass and radius, while particle physicists may condense hundreds of millions of dimensions from detector readouts into a one-dimensional histogram of reconstructed particle energy. I will demonstrate that such dramatic data reduction makes it mathematically impossible to capture all the relevant information necessary for optimal statistical inference in either domain, potentially also leading to the mismodelling of systematic uncertainties. I design data analysis strategies to perform statistical inference directly on high-dimensional data, enabled by powerful uncertainty quantification and propagation tools. We are now integrating these techniques into open-source statistics software and deploying them on DOE supercomputers as a service, to enable their use in numerous applications across particle physics experiments, and potentially also in astrophysics.

Similarly, a significant challenge in theoretical physics is the vast space of mathematical symmetries available to describe our Universe. Despite the dedicated efforts of theorists to explore this expanse, an overwhelming majority remains uncharted. I led the design of a systematic framework for model building in neutrino physics, leveraging newly available computational and AI tools to uncover new avenues for neutrino theory model building.

Bio: Dr. Aishik Ghosh is a postdoctoral scholar at UC Irvine and an affiliate at Berkeley Lab, focusing on the development of high-dimensional statistical inference and uncertainty quantification methods using AI for particle and astrophysics. He has papers in physics and astrophysics journals, as well at NeurIPS. He also developed the first deep generative models for fast simulation to be deployed in a particle physics experiment in 2018. Recently, Aishik has been developing advanced symbolic regression and reinforcement learning methods to address challenges in theoretical neutrino physics, collaborating with computer scientists at Georgia Tech. Previously, he earned his PhD in particle physics from the University of Paris-Saclay.