Gravitational Waves

Attention-boosted Waveform Reconstruction neural network for binary black holes

Neural network for sky localization of gravitational waves, including pre-merger localization of binary neutron star mergers

We have developed a deep learning model, AWaRe, that is capable of producing accurate reconstructions of gravitational wave signals from real LIGO data. We demonstrate that the model is robust against complex waveform systematics like precession and the presence of higher multipoles.

We have applied the deep learning model, GW-SkyLocator, to demonstrate for the first time, that gravitational wave signals from binray neutron star mergers can be localized before merger.

We have developed a deep learning model, called GW-SkyLocator, to obtain the posterior distribution of the sky location of gravitational wave sources. This method is applicable for all three categories of compact binary coalescences - binary black holes, binary neutron stars and neutron star-black hole binaries.

A deep learning-based workflow for denoising and estimating black hole masses from gravitational wave data.

Deep learning can be used to drastically decrease the processing time of parameter estimation for coalescing binaries of compact …

We demonstrate that a denoising autencoder model, consisting of a CNN encoder and a LSTM decoder can be used to extract binary black hole gravitational wave signals from Gaussian and real LIGO noise with high fidelity.

Machine learning has emerged as a popular and powerful approach for solving problems in astrophysics. We review applications of machine …