Astronomy today is fundamentally different than it was even just a decade ago. Our increasing ability to collect a large amount of data from ever more powerful instruments has enabled many new opportunities. However, such opportunities also come with new challenges. The bottleneck stems from the fact that most astronomical observations are inherently high-dimensional — from “imaging” the Universe at the finest details to fully characterising tens of millions of spectra consisting of tens of thousands of wavelength pixels. In this regime, classical astrostatistics approaches struggle.
I will present two different machine learning approaches to quantify complex systems in astronomy. (1) Reductionist approach: I will discuss how machine learning can optimally compress information and extract higher-order moment information in stochastic processes. (2) A generative approach: I will discuss how generative models, such as normalising flow, allow us to properly model the vast astronomy data set, enabling the study of complex astronomy systems directly in their raw dimensional space.
Presented by the Department of Astronomy & Astrophysics.