Applications of probabilistic deep learning for gravitational wave parameter inferenceAbstract
In the era of multi-messenger astronomy, it is imperative to conduct rapid estimation of the masses and sky directions of gravitational wave sources for prompt electromagnetic follow-up observations. However, currently used Bayesian parameter estimation techniques by the LIGO-Virgo collaboration are not fast enough to enable detections of electromagnetic signatures arising during or just after the merger of the binary system. In this talk, I gave an overview of the efforts conducted using probabilistic deep learning algorithms for fast likelihood-free inference of gravitational wave source parameters at comparable accuracy to the optimal Bayesian inference results. I also discussed deep learning techniques for extraction of binary black hole gravitational wave signals from noise and applications of the method for accurate sky localization and chirp mass estimation.
Dr. Chayan Chatterjee
AI for New Messengers Postdoctoral Fellow
Dr. Chayan Chatterjee is the A.I. for New Messengers Postdoctoral Fellow at Vanderbilt University. His research focuses on application of machine learning to study gravitational waves - tiny ripples in space-time created by colliding black holes and neutron stars.