Generalised gravitational wave burst generation with generative adversarial networks

We introduce the use of conditional generative adversarial networks forgeneralised gravitational wave burst generation in time domain.Generativeadversarial are machine learning models that produce new databased on features training data set. condition network fiveclasses time-series signals often used to characterise waveburst searches: sine-Gaussian, ringdown, white noise burst, Gaussian pulse and binaryblack hole merger. show model can replicate these standardsignal classes and, addition, generalised through interpolationand class mixing. also present an example application where a convolutional neuralnetwork classifier is trained generated by our generativeadversarial network. neural trainedonly standard five signal has poorer detection efficiency than aconvolutional population burstsignals drawn from combined space.