David Ginsbourger

Kriging: from geostatistics to machine learning

Seminar stochastics at the University of Bern, weekly sessions starting on February 24th 2011

Current version of the agenda

Session 1: introductory lecture (February 24th)

Beamer presentation on Kriging, from geostatistics to machine learning.

Slides session 1

Slides session 1 (version without pauses)

Session 2: mathematical background, first part (March 3rd)

Blackboard presentation on Gaussian vectors and prediction.

Handouts session 2

Session 3: mathematical background, second part (March 17th)

Blackboard presentation by Ilya Molchanov on spectral decompositions.

Handouts session 3

Session 4: from the linear model to GP regression (April 7th)

Beamer presentation on Gaussian Process Regression, by Brigitte Tschanz.

Slides session 4

Session 5: application of GP in Point Process Intensity Estimation (April 14th)

Blackboard presentation by Dominic Schuhmacher.

Handouts not available yet

Session 6: Basics of GP Classification and application(s) (April 21st)

Beamer presentation by Niki Zumbrunnen.

Slides session 6

Session 7: approximation methods for large datasets (April 28th)

Beamer presentation by Clément Chevalier.

Slides not available yet

Session 8: covariance functions and model selection (Mai 5th)

Beamer presentation by Reinhard Hanselka Jr.

Slides not available yet