[Skip to Content]

Supervisors: Kevin Vinsen, Chen Wu: chen.wu@icrar.org, Andreas Wicenec, Linqing Wen, Richard Dodson: richard.dodson@icrar.org


This PhD will join a fast developing research area for DIA. Currently, Gravitational Waves (GW) are detected from passing the optical interferometric signals of the LIGO and VIRGO detectors through a matched-filter library of predicted responses. This has long been held to provide the optimal signal to noise for any detection. Our recent work shows that we can surpass the performance of the matched filters, through a machine-learning convolution network. The network is able to utilise the behaviour of the full signal; the relationship between different instances that track the build-up of the signal until the moment of the Black Hole coalescence. The approach is surpassing the performance of matched filters by a factor of 3, increasing the volume of space to which we are sensitive by a factor of 30. Many new GW events can be expected to be discovered in both historical and coming datasets. The PhD student will be involved in the implementation of this ML search algorithm on the LIGO network.


There are many industrial applications of this project; the student will have the opportunity to be embedded with a number of our industrial partners to apply this approach to current industry problems.