🗓️ Week 08 - Gaussian Processes

Author

Affiliation

Dr. Simone Rossi

EURECOM

This week we introduce Gaussian Processes (GPs) as a flexible, non-parametric approach to Bayesian regression. We will discuss how GPs define a prior over functions, the role of kernel functions in encoding assumptions about smoothness and structure, and how posterior inference yields both predictions and calibrated uncertainty estimates.

🖥️ Lecture Slides

• Gaussian Processes

📑 Lecture Notes (PDF)

• Gaussian Processes

📝 Exercises

• Gaussian processes exercises

📚 Recommended Reading

• Murphy (2023): Chapter 17.1, 17.2 (Gaussian Processes)

📙 References

Murphy, K. P. (2023). Probabilistic machine learning: Advanced topics. MIT Press. http://probml.github.io/book2