Learning to Compare Examples

NIPS'06 Workshop

December 8, 2006, Whistler, BC, Canada
Overview - Program - People - References

Overview

The identification of an effective function to compare examples is essential to several machine learning problems. For instance, retrieval systems entirely depend on such a function to rank the documents with respect to their estimated similarity to the submitted query. Another example is kernel-based algorithms which heavily rely on the choice of an appropriate kernel function. In most cases, the choice of the comparison function (also called, depending on the context and its mathematical properties, distance metric, similarity measure, kernel function or matching measure) is done a-priori, relying on some knowledge/assumptions specific to the task. An alternative to this a-priori selection is to learn a suitable function relying on a set of examples and some of its desired properties. This workshop is aimed at bringing together researchers interested in such a task.

Topics of interest include, but are not limited to:

algorithmic approaches for distance metric(*) learning,
comparisons of distance metric learning(*) approaches,
effect of distance metric(*) learning on retrieval/categorization models,
learning a distance(*) robust to certain transformations,
links between distance(*) learning and the ranking/categorization problem,
criteria, loss bounds for distance(*) learning,
using unlabeled data for distance(*) learning,
applications of the above to IR/categorization problems for text, vision...

(*) "distance metric" can, of course, be replaced by similarity measure, kernel or matching measure as mentioned in the abstract.

Program

Friday, December 8th 2006
Morning Session: 7:30am - 10:30am

7:30am	Introduction Learning to Compare Examples D. Grangier	[slides]
8:00am	Invited Talk Learning Similarity Metrics with Invariance Properties Y. LeCun	[slides]
8:45am	Learning Visual Distance Function for Object Identification from one Example E. Nowak and F. Jurie	[paper] [slides]
9:10am	Coffee break
9:30am	Learning to Compare using Operator-Valued Large-Margin Classifiers A. Maurer	[paper] [slides]
9:55am	Conformal Multi-Instance Kernels M. B. Blaschko and T. Hofmann	[paper] [slides]
10:20am	Discussion Suggested Topic: Applications of Learning to Compare Examples

Afternoon Session: 3:30pm - 6:30pm

3:30pm	Invited Talk Neighbourhood Components Analysis and Metric Learning S. Roweis	[paper] [slides]
4:15pm	Fast Discriminative Component Analysis for Comparing Examples J. Peltonen, J. Goldberger and S. Kaski	[paper] [slides]
4:40pm	Information-Theoretic Metric Learning J. Davis, B. Kulis, S. Sra and I. Dhillon	[paper] [slides]
5:05pm	Coffee break
5:25pm	Statistical Translation, Heat Kernels, and Expected Distances J. Dillon, Y. Mao, G. Lebanon and J. Zhang	[paper] [slides]
5:50pm	Structured Network Learning S. Andrews and T. Jebara	[paper] [slides]
6:15pm	Discussion Suggested Topic: Kernel and Distance Learning

People

Invited Speakers

Yann LeCun, New York University
Sam Roweis, University of Toronto

Organizers

Samy Bengio, IDIAP Research Institute
David Grangier, IDIAP Research Institute

Program Committee

Samy Bengio, IDIAP Research Institute
Gilles Blanchard, Fraunhofer FIRST
Chris Burges, Microsoft Research
François Fleuret, EPFL
David Grangier, IDIAP Research Institute
Thomas Hofmann, Google
Guy Lebanon, Purdue University
Thorsten Joachims, Cornell University
Yoram Singer, The Hebrew University
Alex Smola, National ICT Australia

References

S. Chopra, R. Hadsell and Y. LeCun,
Learning a Similarity Metric Discriminatively, with Application to Face Verification (CVPR, 2005).
K. Crammer, J. Keshet and Y. Singer,
Kernel Design using Boosting (NIPS, 2002).
F. Fleuret and G. Blanchard,
Pattern Recognition from One Example by Chopping (NIPS, 2005).
A. Globerson and S. Roweis,
Metric Learning by Collapsing Classes (NIPS, 2005).
J. Goldberger, S. Roweis, G. Hinton, R. Salakhutdinov,
Neighbourhood Component Analysis (NIPS, 2004).
D. R. Hardoon, S. Szedmak and J. Shawe-Taylor,
Canonical Correlation Analysis: An Overview with Application to Learning Methods (Neural Comp., 2004).
T. Hertz, A. Bar-Hillel and D. Weinshall,
Boosting Margin-Based Distance Functions for Clustering (ICML, 2004).
G. R. G. Lanckriet, N. Cristianini, P. Bartlett, L. El Ghaoui and M. I. Jordan,
Learning the Kernel Matrix with Semidefinite Programming (JMLR, 2004).
G. Lebanon,
Metric Learning for Text Documents (TPAMI, 2006).
S. Shalev-Shwartz, Y. Singer and A. Y. Ng,
Online and Batch Learning of Pseudo-Metrics (ICML, 2004).
M. Schutz and T. Joachims,
Learning a Distance Metric from Relative Comparisons (NIPS, 2003).
K. Q. Weinberger, J. Blitzer, and L. K. Saul,
Distance Metric Learning for Large Margin Nearest Neighbor Classification (NIPS, 2005).
E. Xing, A. Y. Ng, M. Jordan, and S. Russell,
Distance Metric Learning, with Application to Clustering with Side-Information (NIPS, 2002).