Speakers
Mr
Bill Atkins
(Swansea University)
Matthew Di Ferrante
(Independent Researcher)
Description
We apply a Convolutional Neural Network (CNN) to Pulsar Timing Array residuals to identify the cosmological contribution to the Stochastic Gravitational wave Background for a variety of different cosmological models.
We find the CNN can accurately identify the cosmological contributions, and reconstruct injected signals with at least as much success as current Bayesian methods, but with considerably greater model coverage.
We apply this to generate bounds on the amplitudes and spectral indexes for different cosmological models required to disentangle the two contributions.
Primary authors
Mr
Bill Atkins
(Swansea University)
Matthew Di Ferrante
(Independent Researcher)
