| Name: | Dileep Pavan Surya Jakka |
| Affiliation: | Indian Institute of Technology Kharagpur |
| Conference ID : | ASI2022_736 |
| Title : | Detection of gravitational waves at LIGO using deep learning |
| Authors : | 1. Dr. Sree Ram Valluri 2. Dr. Ramit Dey |
| Abstract Type: | Poster* |
| Abstract Category : | General Relativity and Cosmology |
| Abstract : | Detection of more than 90 Gravitational wave (GW) merger events of compact binary objects by LIGO opened a new window into multi-messenger astronomy. With this the demand for having a very accurate and fast trigger for GW events detected at LIGO is growing exceedingly. Recently, various machine learning techniques (in particular deep learning algorithms) has been used as a means to achieve this objective with a very high precision. In this work we implemented the use of multi-modal deep learning networks, working with both the timeseries GW data as well as its time-frequency domain representation to built a classifier that can detect the GW signal of binary black hole mergers buried in noise. For the purpose of training the model we used noise obtained from the LIGO data and generated the GW waveforms using the latest IMRPhenomX waveform generator. Our preliminary analysis shows that using the concatenated network we can achieve a higher accuracy of detection compared to using only the timeseries data or the time frequency domain data. We are working towards incorporating the approach of Bayesian deep learning into our model so that we can estimate the confidence of the detection. |