The application of machine learning to astronomical imaging surveys

Lead Research Organisation: Durham University
Department Name: Physics

Abstract

The aim of this project is to apply a range of machine learning and data mining algorithms to aid the classification of objects from current and future astronomical imaging surveys. With the start of the LSST 10 year survey only 4 years away, the need to have robust and reliable tools to select the rarest objects, e.g. distant quasars, low mass stars, rich clusters of galaxes, is pressing. Using the UKIDSS/PanSTARRS and VIKING/KiDS matched NIR and optical surveys as a training set, these techniques will be optimised. The emphasis will be on exploiting the photometric catalogues to address a wide range of science goals from low mass stars to the most distant quasars.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
ST/P006744/1 01/10/2017 30/09/2024
1944971 Studentship ST/P006744/1 01/10/2017 31/12/2021 Aidan Sedgewick
 
Title Reduced UKIDSS-DXS mosaics and catalogues 
Description mosaics constructed from ukidss dxs individual stacks for 4 dxs fields (elais-n1, lockman hole, sa22, xmm-lss), reduced using SWarp. catalogues produced using SExtractor, matched to PanSTARRS mds catalogues and subaru HSC strategic program (only available in EN1 and XM). 
Type Of Material Database/Collection of data 
Year Produced 2020 
Provided To Others? No  
Impact N/A 
 
Description INSPIRE-HEP secondment 
Organisation inSPIRE High Energy Physics (HEP) Information System
Country Global 
Sector Public 
PI Contribution research in (machine learning) models to classify articles for inclusion in INSPIRE-HEP database - classify as 'core', 'non-core', 'reject' suggestions on features to use for future models (when producing a model for production)
Collaborator Contribution use of the inspire database (and all of the knowledge of how it worked) techinical discussions on models data handling strategies
Impact N/A
Start Year 2018