Contribution of AI and deep learning in revolutionizing gravitational wave detection

Abstract

The fusion of cutting-edge computing techniques with physical detection of gravitational waves can be a potent solution for detecting and cleaning gravitational wave data, which further helps us in the identification of potential astrophysical sources. In this review article, we discuss the role of artificial intelligence approaches in the analysis of gravitational wave data. Below, we list both ground-based interferometers (like LIGO, VIRGO, etc.) and pulse timing arrays (like Parkes pulse timing array) as the current technologies used to find gravitational waves, along with their benefits and how they can be used to find different kinds of gravitational waves. We survey all four types of gravitational waves, each requiring a unique approach to both detection and data processing. We have extensively studied the use of deep learning techniques like convolutional neural networks, autoencoders, and LSTMs in the detection and parameter estimation of gravitational waves from various possible sources, including binary neutron star mergers and neutron star-black hole mergers, in detail. The review article also includes a thorough understanding of the noise and glitches in the real-time data of gravitational waves, as well as how the effective use of machine learning and deep learning techniques can be helpful in simulating waveforms and removing noise to quantify results.