Applications of machine learning for gravitational wave astrophysics
Lead Research Organisation:
University of Glasgow
Department Name: School of Physics and Astronomy
Abstract
The aim of the project is to apply and develop machine learning techniques to the data analysis problems associated with gravitational wave astrophysics. The student will initially tackle the compact binary coalescence detection problem and show that a deep neural network can be used to replicate the performance of traditional matched-filtering approaches. This will be followed by an investigation into the feasibility of performing parameter estimation for well modelled gravitational wave signals with the aim of comparing the machine learning with the optimal Bayesian approach. With these tools in hand the student will then extend the new machine learning tools to other modelled gravitational wave signals and detectors (e.g., 3rd generation and space-based). In addition the student will investigate the use of un-supervised (or weakly supervised) machine learning algorithms for the search and characterisation of un-modelled transient gravitational wave signals (bursts).
People |
ORCID iD |
Chris Messenger (Primary Supervisor) | |
Hunter Gabbard (Student) |
Studentship Projects
Project Reference | Relationship | Related To | Start | End | Student Name |
---|---|---|---|---|---|
ST/P006809/1 | 01/10/2017 | 30/09/2024 | |||
2039703 | Studentship | ST/P006809/1 | 01/10/2017 | 31/03/2021 | Hunter Gabbard |
Title | Bayesian parameter estimation using conditional variational autoencoders for gravitational wave astronomy |
Description | We show for the first time that a form of machine learning may be used to extract interesting source parameter information from gravitational wave detections with about a 7 order of magnitude increase in performance over existing techniques. |
Type Of Material | Improvements to research infrastructure |
Year Produced | 2019 |
Provided To Others? | Yes |
Impact | This tool (within the short amount of time it's been available) has already made a large impact on the gravitational wave community. It has been used in numerous follow-up studies and has encouraged many other physicists to pursue this new area of research with alternative methods. |
URL | https://arxiv.org/abs/1909.06296 |
Title | Matching matched template filtering using convolutional neural networks for gravitational wave astronomy |
Description | We show for the first time that a form of machine learning can be used to detect gravitational waves with massive speed-ups in performance over existing techniques. |
Type Of Material | Improvements to research infrastructure |
Year Produced | 2018 |
Provided To Others? | Yes |
Impact | It has been widely used for other follow-up studies by a large fraction of the gravitational wave community. |
URL | https://arxiv.org/abs/1712.06041 |
Description | Eagle Scout Interview for Austin, TX media outlet |
Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Public/other audiences |
Results and Impact | Did an interview for a local Austin, Texas media broadcaster which discussed my career path and how it relates to what I learned as an Eagle scout. Reached an audience probably in the 10s of thousands. |
Year(s) Of Engagement Activity | 2018 |
URL | https://www.youtube.com/watch?v=vGceRTvM4AM |
Description | MarketScale gravitational wave podcast |
Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Media (as a channel to the public) |
Results and Impact | Helped make a few podcast episodes centred around our own work and gravitational waves in general. |
Year(s) Of Engagement Activity | 2018 |
URL | https://marketscale.com/industries/sciences/part-1-nature-learning-hunter-gabbard |
Description | Outreach Skype event for American High School class |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Schools |
Results and Impact | Did a skype interview for a class on the everyday life of an astrophysicist a few times over the course of both 2018 and 2019. Also answered other questions related to gravitational wave astrophysics. |
Year(s) Of Engagement Activity | 2018,2019 |
Description | Physics world piece on our work |
Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Media (as a channel to the public) |
Results and Impact | Did an interview for the physicsworld magazine on a publication we released in 2018. |
Year(s) Of Engagement Activity | 2018 |
URL | https://physicsworld.com/a/artificial-intelligence-spots-gravitational-waves/ |