Machine Learning in Astronomy (IAU S368) : Possibilities and Pitfalls

IAU S368 addresses graduate students and professional astronomers who wish to leverage machine learning to unlock the potential of modern data-rich surveys and deep images, as well as archival data. Researchers at the frontiers share best practices in applied machine learning that are relevant to astronomy and other data-rich fields.

Today''s astronomical observatories are generating more data than ever, from surveys to deep images. Machine learning methods can be a powerful tool to harness the full potential of these new observatories, as well as large archives that have accumulated. However, users should beware of common pitfalls, including bias in data sets and overfitting. IAU Symposium 368 addresses graduate students, teachers and professional astronomers who would like to leverage machine learning to unlock these huge volumes of data. Researchers pushing the frontiers of these methods share best practices in applied machine learning. While this volume is focused on astronomy applications, the methodological insights provided are relevant to all data-rich fields. Machine learning novices and expert users will find and benefit from these fresh new insights.