Machine Learning in the Multi-Messenger Era: Inference as a service and optimal light curve augmentation

Michael Coughlin
University of Minnesota, Twin Cities

In this talk, we present two ways in which machine learning are improving prospects for multi-messenger astronomy. This first is a novel paradigm for the deployment of computing infrastructure for the low-latency analyses of gravitational wave (GW) data. Using “replayed” streams of the GW data of LIGO Hanford and Livingston from their third observing run, we demonstrate the subtraction of stationary and non-stationary noise sources followed by identification of candidates for the astrophysical transient detections at low latency. Depending on the computing platform used, we show that it is possible to achieve these at latencies of ~ a few hundreds of milliseconds to a few seconds. The second is strategic, photometric augmentation of light curves to improve follow-up resource utilization. With the advent of facilities such as the Zwicky Transient Facility, their data throughput has overwhelmed the ability to manually synthesize alerts for devising and coordinating necessary follow-up with limited resources. We present the first implementation of autonomous real-time science-driven follow-up for survey transients.

Presentation (PDF File)

Back to Workshop IV: Big Data in Multi-Messenger Astrophysics