Login

Transferring Nonlinear Representations using Gaussian Processes with a Shared Latent Space

Show full item record




Title: Transferring Nonlinear Representations using Gaussian Processes with a Shared Latent Space
Author: Urtasun, Raquel; Quattoni, Ariadna; Lawrence, Neil; Darrell, Trevor
Other Contributors: Vision
Advisor: Trevor Darrell
Issue Date: 2008-04-11
Abstract: When a series of problems are related, representations derived from learning earlier tasks may be useful in solving later problems. In this paper we propose a novel approach to transfer learning with low-dimensional, non-linear latent spaces. We show how such representations can be jointly learned across multiple tasks in a Gaussian Process framework. When transferred to new tasks with relatively few training examples, learning can be faster and/or more accurate. Experiments on digit recognition and newsgroup classification tasks show significantly improved performance when compared to baseline performance with a representation derived from a semi-supervised learning approach or with a discriminative approach that uses only the target data.
URI: http://hdl.handle.net/1721.1/41517
Other Identifiers: MIT-CSAIL-TR-2008-020
Related To Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory

Files in this item

Files Size Format
MIT-CSAIL-TR-2008-020.pdf 303.1Kb application/pdf
MIT-CSAIL-TR-2008-020.ps 73.87Kb application/postscript

This item appears in the following Collection(s)

Show full item record

Search DSpace@MIT


Advanced Search

Browse

My Account

Links