Skip to main content

Machine learning analysis of gravitational waves

Primary supervisor

David Dowe


  • Paul Lasky
  • Kendall Ackley

The recent discovery in 2015 of gravitational waves from colliding black holes and neutron stars has opened a new window on the Universe. Astrophysicists can now “see the unseeable” -- black holes that emit no light are regularly being observed through their gravitational-wave signatures. Since the first discovery in 2015, more than 50 black hole mergers, two neutron star mergers, and two neutron star-black hole collisions have been observed. This new field of discovery is in its infancy, with many more astrophysical sources waiting to be uncovered such as signals from supernovae explosions and “mountains” on rapidly rotating neutron stars. Perhaps the most exciting future discovery is the first detection of the unknown -- sources not yet predicted by astrophysicists.



Uncovering the unknown is a challenging data problem. Weak signals from collisions of compact objects can be dug out of noisy time series because we understand what the signal should look like, and can therefore use simple algorithms based on matched-filter statistics. Detecting the unknown relies on the development of complex algorithms at the forefront of statistics, machine learning, and data science. This multi-disciplinary project will combine data science and artificial intelligence expertise from the Faculty of Information Technology with gravitational-wave astrophysics expertise from the School of Physics and Astronomy in the Faculty of Science. We will apply Bayesian approaches such as the information-theoretic minimum message length (MML) principle and other approaches to develop a path towards statistically-optimal algorithms to detect unknown signals from gravitational-wave observatories across planet Earth.



 Comley, Joshua W. and D.L. Dowe (2003). General Bayesian Networks and Asymmetric Languages, Proc. 2nd Hawaii International Conference on Statistics and Related Fields, 5-8 June, 2003

 Comley, Joshua W. and D.L. Dowe (2005). ``Minimum Message Length and Generalized Bayesian Nets with Asymmetric Languages'', Chapter 11 (pp265-294) in P. Gru:nwald, I. J. Myung and M. A. Pitt (eds.), Advances in Minimum Description Length: Theory and Applications, M.I.T. Press (MIT Press), April 2005, ISBN 0-262-07262-9. [Final camera ready copy was submitted in October 2003.]

  Wallace, C.S. (2005), ``Statistical and Inductive Inference by Minimum Message Length'', Springer  (Link to the preface [and p vi, also here])

  Wallace, C.S. and D.L. Dowe (1994b), Intrinsic classification by MML - the Snob program. Proc. 7th Australian Joint Conf. on Artificial Intelligence, UNE, Armidale, Australia, November 1994, pp37-44

  Wallace, C.S. and D.L. Dowe (1999a). Minimum Message Length and Kolmogorov Complexity, Computer Journal (special issue on Kolmogorov complexity), Vol. 42, No. 4, pp270-283

 Wallace, C.S. and D.L. Dowe (2000). MML clustering of multi-state, Poisson, von Mises circular and Gaussian distributions, Statistics and Computing, Vol. 10, No. 1, Jan. 2000, pp73-83

Required knowledge

Mathematics at least to 1st year undergraduate level, preferably more.

Statistics, machine learning and/or data science at least to the level of undergraduate degree.

An ability to program.

At least an interest in physics, gravitational waves and better understanding the physics of the universe.

Learn more about minimum entry requirements.