During the last four days I was busy with the actual materialization of the Astrostatistics Summer School Crete 2019, which took place at the University of Crete, in Heraklion (18-21 June 2019). My duties were mainly those of the Teaching Assistant and I contributed with a short presentation of the Random Forests method for classification. Unfortunately I didn’t have the time to post more on this school, so I ended up doing something only today, at the very last day!
In summary this is/was a school for graduate and early-stage post-docs to get a grasp of the modern field of Astrostatistics, which practically means the application of statistics in Astronomy which incorporates also machine-learning techniques. Topics include: Intro to Python and Jupyter notebook, linear regression, classical statistical distribution tests and hypothesis testing, Bayesian statistics, Markov-Chain Monte Carlo, machine learning classification/regression/clustering, time series analysis.
The school is split in “teaching”/explanatory parts (through Jupyter notebooks though where you could also interact and run the examples) and practical workshops, where important hands-on experience with all the tools presented was provided (and hopefully gained!). However, I think the importance lies in the fact that all the material is publicly available through a github repository: https://github.com/astrostatistics-in-crete (including the notebooks with the introduction notes, the exercises in the workshops, and the data). So this is a valuable source both for students of the school, as well as others interested to try and experiment with these tools.