Machine Learning for Exoplanet Discovery


Over recent years our knowledge of planets beyond our Solar System, known as exoplanets, has grown rapidly. We have now detected over 4000 other planets, with many strange and unusual examples unlike anything previously expected. Multiple missions led by NASA, ESA and others are coming online, aiming to investigate these planets in detail and answer the question of how common, or unique, our own Earth and Solar System truly are. More data is available than ever before, requiring rigorous, fast, and statistically defensible algorithms to fully utilise it. I will describe how new exoplanets are discovered, the types of data available, and how machine learning is being used to improve the discovery process. In particular, I will focus on how probabilistic models can be used to identify the true planets in a dataset statistically, cleaning many types of false positive signal. Removing much of the human element from the process allows us to identify true exoplanets reliably and quickly, optimising the use of expensive follow-up facilities and increasing the rate and quality of the resulting scientific discoveries. Machine learning has only recently begun to be applied to the exoplanet field, with many areas still open for exploration.


David Armstrong is an STFC Ernest Rutherford Fellow working on exoplanet detection, characterization and populations with the TESS mission data. He develops and use automated methods of planet detection to improve planet population statistics, both in general and with a past focus on circumbinary planets. David's other interests include studying the habitability of known exoplanets, time-variability in planetary phase curves, and the identification and classification of eclipsing binaries and variable stars in large-scale surveys.

One of his particular interests is in applying machine learning techniques to astrophysical problems. Recently this has included the automatic selection of real candidates in transiting planet surveys, as well as creating fast detection and classification tools for eclipsing binaries and variable stars in the K2 survey. David works with the Computer Science and Statistics departments at Warwick, with the aim of applying the latest machine learning techniques to problems in the exoplanet field.

David Armstrong is currently an academic co-lead of Warwick's Habitability GRP, a new initiative at Warwick aimed at promoting multi-disciplinary research in habitability, and he is involved with the Centre for Exoplanets and Habitability. Our current projects include analyses of galactic habitability, dynamical and biological constraints on panspermia, and the stability of DNA in non-Earth environments.

Open Data Science




Open Data Science
One Broadway
Cambridge, MA 02142

Privacy Settings
We use cookies to enhance your experience while using our website. If you are using our Services via a browser you can restrict, block or remove cookies through your web browser settings. We also use content and scripts from third parties that may use tracking technologies. You can selectively provide your consent below to allow such third party embeds. For complete information about the cookies we use, data we collect and how we process them, please check our Privacy Policy
Consent to display content from Youtube
Consent to display content from Vimeo
Google Maps
Consent to display content from Google