Laboratory of Computer and Information Science / Neural Networks Research Centre CIS Lab Helsinki University of Technology

Courses in previous years: [ 1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 | 2003 | 2004 | Not organized in 2005 ]


T-61.6010 Special Course in Computer and Information Science I L:

Gaussian Processes for Machine Learning 6 cr


News 20.11.2006
  • Equation 4.29 missed an apostrophe. A new version of the second round problems is available.
News 13.11.2006
  • The project assignment has been updated and the second round of exercises published.
News 16.10.2006
  • The exercise assignment paper had a wrong equation in the 8th assignment. A new version of the paper is available.

[ Schedule | Exercise problems | Project work ]


Lecturer D.Sc. Antti Honkela
Assistant M.Sc. Sami Hanhijärvi
Credits (ECTS) 6
Semester Autumn 2006 (during periods I and II)
Seminar sessions On Mondays at 14-16 in lecture hall T4 in computer science building,
Konemiehentie 2, Otaniemi, Espoo. The first session is on September 11, 2006.
Language English
Web http://www.cis.hut.fi/Opinnot/T-61.6010/
E-mail t616010 (at) cis.hut.fi

Introduction

Gaussian processes provide a principled, practical, probabilistic approach to learning in kernel machines. Much of the recent work on kernel methods for support vector machines (SVMs) is directly applicable to Gaussian processes.

This course is based on a recent textbook Gaussian Processes for Machine Learning by Carl E. Rasmussen and Christopher K. I. Williams. The book presents a comprehensive treatment of Gaussian processes for supervised learning problems of regression and classification including detailed algorithms. Different covariance (kernel) functions, model comparison issues, connections to related models and approximations for learning with large datasets are also discussed. Additional topics on Gaussian processes in unsupervised learning will also be considered based on additional material.

Prerequisites

Basic knowledge on pattern recognition (e.g. T-61.3020 Principles of Pattern Recognition) and Bayesian inference methods (e.g. T-61.5040 Learning Models and Methods) will be helpful.

Course format

Seminar course

Requirements for passing the course

Each student gives a presentation in the seminar. In addition, requirements include solving sufficiently many exercise problems and making a project work, as well as active participation in the lectures (one absence is allowed).

Course material

The course mostly follows the book Gaussian Processes for Machine Learning by Carl E. Rasmussen and Christopher K. I. Williams (MIT Press, 2006). In addition to the book, some other material will be discussed.


[ Schedule | Exercise problems | Project work ]


For more information, please send email to the course organisers (t616010 (at) cis.hut.fi).

You are at: CIS → T-61.6010 Special Course in Computer and Information Science I

Page maintained by t616010 (at) cis.hut.fi, last updated Saturday, 01-Sep-2007 09:23:03 EEST