Machine Learning and Artificial Intelligence applied to Astronomy
Data sets in astronomy are becoming extremely large and complex, the research questions that are being asked of these data are also becoming complex and in many cases, the richness of the data surpasses the level of sophistication of the theoretical models. Machine learning and AI can thus be used to augment physical models for practical applications (e.g. photometric redshifts) or physical understanding (e.g. galaxy classification, model fitting).
Meanwhile, the UK government has recently identified artificial intelligence as a key opportunity (one of its four grand challenges) in the Industrial Strategy White paper. As a highly computationally literate community with extreme data challenges, astronomy research could provide a very valuable environment for developing the skills and techniques the UK economy needs. This would provide a significant socio-economic impact, important for the sustainability of the discipline.
Details of the meeting agenda are in the attached PDF document.
Robert Schumann - SNMachine: automated photometric transient classification for LSST
Daniel Muthukrishna - Real-time classification of transients using deep Recurrent Neural Networks
Andrew Davies - Strong Gravitational Lens detection with Machine Learning
Honhming Tang - Transfer learning for radio galaxy classification
Joe Hanson - Investigating cosmic magnetism with deep learning