GW9 - Advanced Data-Analysis Techniques for Gravitational-Wave Detection |
|
Speaker |
Razzano, Massimiliano |
|
Coauthors |
M. Razzano & E. Cuoco |
| Talk Title |
Deep learning for the study of noise in gravitational wave interferometers |
|
Abstract |
The detection of gravitational waves has opened new avenues for the gravitational wave astronomy and the multimessenger study of the cosmos. Thanks to their sensitivity, the Advanced LIGO and Advanced Virgo interferometers will probe a much larger volume of space and discover new sources of gravitational waves. The characterization of the inteferometers is very important in order to recognize the main sources of noise, optimize the sensitivity, and provide a low-latency assessment of data quality. Instrumental glitches are transient noise events that impact the data quality of the interferometers, and their classification is very important to characterize the detector. Deep learning techniques are a promising tool to recognize and classify these glitches. We will discuss the potential of deep learning in studying noise in gravitational wave detectors and distinguishing it from real astrophysical signals. In particular, we will present a method based on deep learning for the classification of glitches based on their time-frequency evolution represented as images. Using tests on simulated data, we show its capability of disentangling the noise from real gravitational signals, and its accuracy in classification coupled with short execution times, thus providing a promising tool for online, low-latency, detector characterization. |
|
Talk view |
|
|
|
|