Connectionists: NIPS*06 Workshop on Learning with Different Input Distributions

[Joaquin Quiñonero Candela]

CALL FOR ABSTRACTS

======================================================
NIPS*06 Workshop - Whistler, BC, December 8-9 2006

"Learning when Test and Training Inputs Have Different
Distributions"

http://ida.first.fraunhofer.de/projects/different06
======================================================

Call for contributions:
---
We invite submissions of extended abstracts (1 to 4 pages
long). A selection of the submitted abstracts will be
accepted as oral presentations. We will also accept a few
abstracts for poster presentation. Oral presenters are
strongly encourage to additionally prepare a poster. The
best abstracts will be considered for extended versions for
the workshop proceedings. (please see workshop website for
information about how to submit)

Important Dates:
---
. deadline for submissions: November 8, 2006
. notification of acceptance: November 15, 2006

Program Committee:
---
. Tony O'Hagan (University of Sheffield)
. Bernhard Schoelkopf (Max Planck Institute for
                          Biological Cybernetics)
. Thorsten Joachims (Cornell University)

Background:
---
Many machine learning algorithms assume that the training
and the test data are drawn from the same distribution.
Indeed many of the proofs of statistical consistency, etc.,
rely on this assumption. However, in practice we are very
often faced with the situation where the training and the
test data both follow the same conditional distribution,
p(y|x), but the input distributions, p(x), differ. For
example, principles of experimental design dictate that
training data is acquired in a specific manner that bears
little resemblance to the way the test inputs may later be
generated.

The open question is what to do when training and test
inputs have different distributions. In statistics the
inputs are often treated as ancillary variables. Therefore
even when the test inputs come from a different
distribution than the training, a statistician would
continue doing ``business as usual''. Since the conditional
distribution p(y|x) is the only one being modelled, the
input distribution is simply irrelevant. In contrast, in
machine learning the different test input distribution is
often explicitly taken into account. An example is
semi-supervised learning, where the unlabeled inputs can be
used for learning. These unlabeled inputs can of course be
the test. Additionally, it has recently proposed to
re-weight the training examples that fall in areas of high
test input density for learning (Sugiyama and Mueller,
2005).  Transductive learning, which concentrates the
modelling at the test inputs, and the problem of unbalanced
class labels in classification, particularly where this
imbalance is different in the training and in the test
sets, are both also very intimately related to the topic of
this workshop.

It does not seem to be completely clear, whether the
benefits of explicitly accounting for the difference
between training and test input distributions outweigh the
potential dangers. By focusing more on the training
examples in areas of high test input density, one is
effectively throwing away training data. Semi-supervised
learning on the other hand is very dependent on certain
prior assumptions being true, such as the cluster
assumption for classification.

The aim of this workshop will be to try and shed light on
the kind of situations where explicitly addressing the
difference in the input distributions is beneficial, and on
what the most sensible ways of doing this are.

Organizers:
---
. Joaquin Quinonero Candela (Technical University of Berlin)
. Neil D. Lawrence (University of Sheffield)
. Anton Schwaighofer (Fraunhofer FIRST.IDA)
. Masashi Sugiyama (Tokio Institute of Technology)