Prev: P1.1 Next: P1.3

P1.2: Dreissigacker, Christoph
Christoph Dreissigacker (Albert Einstein Institute Hannover)
Rahul Sharma (Albert Einstein Institute Hannover)
Reinhard Prix (Albert Einstein Institute Hannover)
Christopher Messenger (SUPA, School of Physics & Astronomy, University of Glasgow)




Theme: Machine Learning in Astronomy
Title: Deep-Learning Continuous Gravitational Waves

The search for continuous gravitational waves from unknown spinning neutron stars presents an open computational challenge: optimal fully-coherent matched filtering is computationally impossible, and empirical semi-coherent methods are the best current alternative known. There has been promising progress recently in applying Deep Convolutional Neural Networks as a detection method for binary black-hole coalescence signals (George&Huerta, Gabbard et al, Gebhard et al (2017)). Here we present results of our study on the feasibility and potential of using similar networks to search for continuous gravitational waves directly in the detector strain data.

Link to PDF (may not be available yet): P1-2.pdf