Gaussian processes confer a Bayesian nonparametric framework to model time series data or general one-dimensional data and have recently demonstrated modelling success across a wide range of spatial and temporal application domains.
In the context of astrophysics, there is a recent trend favouring non-parametric models such as Gaussian processes due to the flexibility afforded when specifying the underlying data modelling assumptions. Applications have arisen in modelling stellar activity signals in radial velocity data, learning imbalances for variable star classification, exoplanet detection, spectral modelling as well as AGN variability studies.
These two-day online-only workshops hosted by the Institute of Astronomy, University of Cambridge aim to bring together astronomers and machine learners to discuss the application of Gaussian Processes in astronomical data analyses. We also welcome astronomers with no Gaussian Process background to attend our workshops.
Our workshops will include the Astronomy session on Day 1 and the Machine Learning session on Day 2. Invited Machine Learning experts from both academia and industry will prepare the Machine Learning session according to the submitted Astronomy abstracts. We also accept a small number of contributed Machine Learning talks.
Important Dates
23 September 2022 Abstract Submission Deadline for Astronomy Talk
4 October 2022 Release of the Astronomy Abstract Book
4 October 2022 Release of the Astronomy/Day 1 talk schedule
20 October 2022 Abstract Submission Deadline for Machine Learning Talks
28 October 2022 Release of the Machine Learning/Day 2 talk schedule
4 November 2022 Registration Deadline
14 November 2022 (Mon) Workshop Day 1
21 November 2022 (Mon) Workshop Day 2
Platform: Zoom (The Zoom link will be sent via emails.)
Schedule
14 Nov, 2022
15:00-15:40 (UTC) Dahai Yan (CAS, China)
Gaussian Process Modeling Gamma-ray Variability of AGNs
15:40-16:20 Collin Lewin (MIT, US)
Mapping Out AGN Accretion Disks using Gaussian Processes
16:20-17:00 Dan Wilkins (Standford, US)
Extending X-ray Reverberation Mapping beyond the Local Supermassive Black Holes with Gaussian Processes
21 Nov 2022
15:00-15:40 Fergus Simpson (Secondmind)
Incorporating Domain Knowledge into Gaussian Processes
15:40-16:20 Henry Moss (Secondmind)
Title: TBC
16:20-17:00 Ryan Giffiths (Meta)
Modelling the Multiwavelength Variability of Mrk 335 using Gaussian Processes
17:00-17:40 William Alston (ESAC, Spain)
Simulating Accreting Black Hole Light Curves using Neural Networks
Register & Submit an Abstract (for astronomers)
Deadline: 23 September
If you wish to give a talk, please submit your abstract via the link. You don't need to take any further steps to register for the workshops.
If you do not wish to give a talk but still would like to join the workshops, please register using the link too. There will be no registration fee.
We will compile all submitted abstracts and release the Astronomy Abstract Book on 4 October. This book will include both accepted and rejected abstracts and enable our speakers to better prepare their Machine Learning talks accordingly. Rejected abstracts are also included, so you may have a chance to discuss your project with an expert during or after our workshop.
At the end of the abstract submission form, you will be asked for permission to publish your abstract as part of the Astronomy Abstract Book. If you do not wish to publish your abstract in the Astronomy Abstract Book, please indicate so in the registration form.
Register & Submit an Abstract (for machine learners)
Deadline: 20 October
The Machine Learning session on Day 2 will be mostly invited talks. But we also accept abstract submission for Day 2. A small number of contributed talks will be selected.
If you wish to give a talk, please submit your abstract via the link. You don't need to take any further steps to register for the workshops.
If you do not wish to give a talk, please register using the link too. There will be no registration fee.
We encourage you to read the Astronomy Abstract Book before preparing your talk. The abstract book will be released on 4 October.
Committee
Jiachen Jiang (Uni. of Cambridge)
Henry Moss (Secondmind)
Ryan-Rhys Griffiths (Facebook)