280522 VU Data Science in Astrophysics (2024S)

A big part of modern astrophysics and astronomy consists in working with large datasets obtained from observatories and supercomputing facilities. This lab course will cover essential aspects of modern statistics, data analysis, and machine learning with a focus to the application to astrophysical problems. The courses will include introductory lectures, followed by hands-on sessions using astrophysical data sets (where possible). The relevant astrophysical concepts will be introduced as well, so that for non-astrophysics majors a key interest will lie in the application of modern methods to relevant astrophysical problems.

Topics will include (sometimes split into multiple units, and not necessarily in chronological order):
- density estimation
- spatial statistics
- data representation and compression, singular value decomposition, autoencoders
- regression
- gaussian processes
- classical machine learning
- supervised machine learning and neural networks
- physics informed neural networks
- generative models
- data visualisation

The final part of the course is dedicated to a ‘focus project’, i.e. an individually carried out small project extending an aspect of the course beyond what is covered.