idamjanov / kavli2019

Lectures for the 2019 Kavli Summer Program in Astrophysics

Kavli Summer Program in Astrophysics Lectures

Lectures by David Kirkby on machine learning for the Kavli Summer Program in Astrophysics hosted by UC Santa Cruz in July 2019:

• Practical Bayesian Inference [notebook]
• Unsupervised Deep Learning [notebook]

The slides are based on jupyter notebooks that contain all of the code used for the examples and plots (usually in cells that the slideshow skips). Use the [notebook] links above to view these notebooks. To run the notebooks yourself, you can download them and create a conda environment with the necessary packages:

git clone https://github.com/dkirkby/kavli2019
cd kavli2019
conda env create -f environment.yml
conda activate K19
jupyter notebook

These lectures use material from the UC Irvine Machine Learning & Statistics for Physicists grad course.

If you are interested in presenting your own notebooks as a slideshow, see here for an overview. There are two important caveats (as of July 2019):

• The slideshow feature currently only works with the older jupyter notebook front end, not the new jupyter lab.
• You cannot reliably edit or run cells from within the slideshow (the RISE package is designed for this, but was not reliable for me).

If find errors or have suggestions for improvement, please create an issue.

This material is (c) 2019 David Kirkby [e-mail] and released under an MIT License.