This repository contains tutorials for the 2023 KITP hybrid program Building a Physical Understanding of Galaxy Evolution with Data-driven Astronomy. For the program schedule and additional information, take a look at our website or KITP listing.

The tutorial sessions will cover various topics in galaxy evolution and scientific machine learning. Some are designed to provide an introduction or overview of statistical and machine learning methods. Others focus on machine learning applications or answering specific scientific questions.

Although each tutorial session is 1.5 hours long, we expect that the tutorial leads should keep their presentation to under an hour, which allows attendees to ask questions, work interactively, or discuss added topics.

Each week we will address a different set of scientific topics, and these weekly themes are now set for the virtual programs (see here). We aim to have two tutorials per week, and will update this page as more details are finalized.

• Week 1 Tutorials
  • Wed 1/18 - [Colab] [Recording] Large-scale galaxy formation simulations and machine learning approaches, part 1: Absorption spectra in hydro simulations, by Mahdi Qezlou
  • Wed 1/18 - [Colab] [Recording] Large-scale galaxy formation simulations and machine learning approaches, part 2: Gaussian process regression for emulation, by Ming-Feng Ho
  • Thurs 1/19 - [Colab] [Recording] Introduction to convolutional neural networks, by John Wu
• Week 2 Tutorials
  • Tues 1/24 - [Colab] [Recording] The galaxy-halo connection and machine learning approaches, part 1: Modeling the halo-galaxy connection, by Natalí de Santi
  • Tues 1/24 - [Colab] [Recording] The galaxy-halo connection and machine learning approaches, part 2: Density estimation with normalizing flows, by Christopher Lovell
  • Wed 1/25 - [Colab] [Recording] Galaxy scaling relations through working with UniverseMachine, by Peter Behroozi
• Week 3 Tutorials (CCA hybrid workshop)
  • Mon 1/30 - [Colab] Robust Uncertainty Estimation in Machine Learning, by Aritra Ghosh
  • Tues 1/31 - [Github] Symbolic Regression with PySR, by Miles Cranmer
  • Wed 2/1 - [Colab] Simulation-Based Inference, by ChangHoon Hahn
  • Thurs 2/2 - [Github] Short tutorial on PyCaret, by Toby Brown
  • Thurs 2/2 - Spectral energy distribution modeling, by Kartheik Iyer
  • Fri 2/3 - [Colab] Graph neural networks and merger trees, by Christian Kragh Jespersen
• Week 4 Tutorials
  • Wed 2/8 - [Colab] Photometric redshifts and uncertainties thereof, by Alex Malz
• Week 5 Tutorials
  • Wed 2/15 - [Github] Autodifferentiation and JAX, by Andrew Hearin
• Week 6 Tutorials
  • Wed 2/22 - [Colab] Probabilistic U-Nets, by Hadi Sotoudeh

We hope that most of these tutorials can be run on the cloud (e.g. through Google Colab) or locally with minimal dependencies. However, we note that different tutorials have different authors, so the coding and writing style will not be consistent throughout the repository.

To navigate to a tutorial, simply click on the week you're interested in (see above schedule) and find the relevant files for each tutorial. For example, if you want to open the Introduction to Convolutional Neural Networks tutorial, go to week-1 and then open Introduction to convolutional neural networks.ipynb. Note that you can also open this repository or the subdirectories in Google Colab.