Frontiers at the interface of deep learning and large scale computing for Multi-Messenger Astrophysics
Abstract
We discuss recent developments at the interface of deep learning and large scale computing for the design and use of deep neural network models for the detection of gravitational waves in non-Gaussian and non-stationary noise, and the characterization of the binary components and the remnant of black hole mergers. We discuss the generality of these methodologies for the analysis of telescope image datasets, and showcase its application with the construction of large scale galaxy catalogs with the Dark Energy Survey. We also present neural network models, endowed with a statistical backbone, for gravitational wave parameter estimation analyses.
- Publication:
-
APS April Meeting Abstracts
- Pub Date:
- 2019
- Bibcode:
- 2019APS..APRD17009H