University of Edinburgh
School of Mathematics and BioSS
Date: Friday 11th Nov, 15:10pm Location: JCMB 5327
Speaker: Dr Michael Gutmann, School of Informatics,
University of Edinburgh
Title: Noise-contrastive estimation of unnormalised statistical models
Abstract: Parametric statistical models are often not properly normalised, that
is, they do not integrate to unity. While unnormalised models can, in
principle, be normalised by dividing them by their integral, the cost of
computing the integral is generally prohibitively large. This is an
issue because without normalisation, the likelihood function is not
available for performing inference.
I present a method called "noise-contrastive estimation" where
unnormalised models are estimated by solving a classification problem. I
explain some of its properties and applications, and show that it is
part of a general estimation framework based on the Bregman divergence.
Related papers:
http://jmlr.csail.mit.edu/papers/v13/gutmann12a.html
http://arxiv.org/abs/1202.3727
--
There will be tea and coffee after the seminar in the Mathematics Common Room.
This seminar is a part of Maxwell Institute seminar series.
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|