Ay 119: Astroinformatics - Spring 2025

Open to everyone interested - students from any field are welcome.

Instructors:

• George Djorgovski [djorgovski]
• Ashish Mahabal [ashish]
• Matthew Graham [mjg]
• Santiago Lombeda [santiago]
• [name] at caltech dot edu

---

Syllabus (pdf)

Lectures / Discussion sessions: Tuesdays, 4-5 pm in 304 Cahill

Schedule by the week (subject to changes):

• Week 1 - George: Introduction
  • Assigned reading:
    • Dalya Baron: "Machine Learning in Astronomy: a practical overview" (arXiv:1904.07248), also this DOI
    • S. G. Djorgovski, A. A. Mahabal, M. J. Graham, K. Polsterer, A. Krone-Martins: "Applications of AI in Astronomy" (arXiv:2212.01493), also this DOI
  • Supplementary viewing: George's lecture about the Astronomy Data Landscape, from the KISS short course, May 2019, and his slides (pdf)

• Week 2 - George/Ashish/Matthew: ML Foundations
  • Umaa Rebbapragada' ML Basics from AstroInformatics 2019: Video [starts at 1:15] * Slides (pdf)
  • Ciro Donalek's Introduction to ML from Caltech-JPL Virtual Summer School in Big Data Analytics: Video * Slides (pdf)

• Week 3 - Ashish: Supervised Learning

• Week 4 - Ashish: Unsupervised Learning: Clustering, Dimensionality Reduction

• Week 5 - Santiago: Data Visualization

• Week 6 - Ashish/Matthew: Large Language Models

• Week 7 - Matthew: Introduction to Deep Learning: Neural Networks, CNNs, RNNs, Autoencoders

• Week 8 - Matthew: Advanced Deep Learning: Transformers, Intro to Reinforcement Learning and Physics-Informed Neural Networks (PINNs)

Resources and Useful Links:

• There are many, many excellent resources online, including Wikipedia. Let us know if you find some that you really like, and we'll add them here

• International AstroInformatics Association resources page has many useful links (needs updating)

• We highly recommend Prof. Yaser Abu-Mostafa's class Learning From Data (aka CS 156)

• Caltech Libraries provide a free access to the O'Rilley books and videos (Caltech Access login required). This is a great resource.

• Old, but useful lectures from the Caltech-JPL Big Data Analytics Virtual Summer School

• Andrew Moore's ML tutorials from CMU