This tutorial is on Gaussian Processes for Regression: Models, Algorithms, and Inference. It took place at the 7th Computational Physics School for Fusion Research (CPS-FR 2026) and was hosted by the MIT Plasma Science and Fusion Center. It ran on August 5 and 6. See this link for the latest versions of all tutorials.

Part I: Wednesday August 5, 1:30pm–2:30pm
Part II: Thursday August 6, 9:00am–10:00am
Part III: Thursday August 6, 10:15am–11:15am
Part IV: Thursday August 6, 11:30am–12:30pm

The instructor is Prof. Tamara Broderick. Contact information can be found here.

Prerequisites: It will be helpful to have basic familiarity with (1) Bayesian data analysis and its goals and (2) both univariate and multivariate Gaussian distributions though we will review key facts.

Slides:

• Part I: Gaussian process (GP) model [slides for Part I]
• Part II: GP regression and inference [slides for Parts II, III, and IV]
• Part III: Squared exponential kernel and observation noise
• Part IV: Bayesian optimization and (more) limitations of (GP) regression
• See the end of the second set of slides for a list of resources

Code for demos will be posted here:

• Code for demos from demos from all parts (in order) can be found [here]

Errata:

• There are some parts in the code that need cleaning up, but I am not aware of errors in the parts of the code I wrote. Please let me know if you find any though!

Accessibility
Plain Academic