Hide menu

Polopoly will be shut down December 15, 2023. Existing pages will have to be moved or will be removed at that date. Empolyees may read more at FAQ Polopoly Avveckling

Gaussian Processes for Machine Learning

Number of credits: 9 hp

Examiner: Timo Koski

Course literature: Gaussian Processes for Machine Learning, Carl Edward Rasmussen and Christopher K. I. Williams MIT Press, 2006.

Course contents:
1 Introduction
1.1 Introduction to Bayesian Modelling
2 Regression
3 Classification
Multi-class Laplace Approximation
4 Covariance Functions
5 Model Selection and Adaptation of Hyperparameters
6 Relationships between GPs and Other Models
6.1 Reproducing Kernel Hilbert Spaces
6.2 Regularization
6.3 Spline Models
6.4 Support Vector Machines
6.5 Least-Squares Classification
6.6 Relevance Vector Machines
7 Theoretical Perspectives
7.1 The Equivalent Kernel
7.2 Asymptotic Analysis
7.3 Average-case Learning Curves
7.4 PAC-Bayesian Analysis
7.5 Comparison with Other Supervised Learning Methods
7.6 Appendix: Learning Curve for the Ornstein-Uhlenbeck Process

Organisation: Seminars and computer exercises.

Examination: Home exam.

Prerequisites:: TAMS47 or TAMS46. En tidigare kurs i statistisk inlärningsteori eller statistisk inferens inte nödvändig.

Page manager: karin.johansson@liu.se
Last updated: 2014-04-29