Call for Abstracts
New Directions in Statistical Learning for Meaningful and Reproducible
NIPS 08 Workshop, Whistler, Canada
* Submission deadline (2 page abstract): October 31
* Notification of acceptance: November 7
* Workshop date: December 13
Over the last several years, statistical learning methods have become
mainstream in the analysis of Functional Magnetic Resonance Imaging
(fMRI) data, spurred on by a growing consensus that meaningful
neuroscientific models built from fMRI data should be capable of
accurate predictions of behavior or neural functioning. Two years ago,
the NIPS workshop "New Directions on Decoding Mental States from fMRI
Data" reflected on progress so far and future directions. Most of the
open questions discussed considered how to advance beyond
single-subject, single-task, voxel-by-voxel, static analysis to better
uncover the true underlying activation patterns and thus better
characterize brain functioning.
Two years later, the field has continued to see great success in
predictive modeling, as the results of the 2006 and 2007 Pittsburgh
Brain Activity Interpretation Competition demonstrate, convincing most
neuroscientists that there is tremendous potential in the decoding of
brain states using statistical learning. Along with this realization,
though, has come a growing recognition of the limitations inherent in
using black box methods for drawing neuroscientific interpretations. The
primary challenge now in the field is how best to exploit statistical
learning to answer scientific questions by incorporating domain
knowledge and embodying hypotheses about various cognitive processes.
Further advances in the field will require resolution of many open
questions, including the following:
* To what extent do patterns in fMRI replicate across trials, subjects,
tasks, and studies?
* To what extent are processes that are observable through the BOLD
response measured by fMRI truly replicable across these different
* How similar is the neural functioning of one subject to another?
* The most common data representation continues to consider voxels as
static and independent, and examples are i.i.d.; however, voxels
represent arbitrary spatial subdivisions of the brain space; hence,
activation patterns almost surely do not lie in voxel space. What are
the true, modular activation structures?
* What is the relationship between similarity in cognitive state space
and similarity in brain state space?
* Brain functioning is clearly a dynamical system, and the fMRI images
indirectly measuring this functioning are not static and independent,
but rather a snapshot in time. To what extent can causality be inferred
This 1-day workshop will serve to engage leaders in the field in a
debate about these issues while providing an opportunity for
presentation of cutting-edge research addressing these questions.
The workshop will begin with a tutorial introduction to the broad area
of statistical learning for fMRI analysis, and will then be divided into
2 sessions roughly corresponding to the 2 topics outlined above, with
each session featuring an overview talk on the issue by a leader in the
field, followed by shorter submitted talks and a panel discussion. The
workshop will conclude with a group discussion on controversies in
generalizability, robustness, data representations, and other topics.
Depending on the number of submissions, we may also have a poster
session for additional submitted abstracts. The target audience will
include both neuroscientists and statistical learning researchers
working with fMRI, as well as a more general audience from both fields.
- Cross-subject / cross-study / cross-task analysis
- Variable selection / dimensionality reduction / sparsity
- Hierarchical models
- Stimulus space representations
- Hypothesis generation and testing / experimental design
- Functional connectivity analysis / network learning
- Dynamic causal modeling
We invite abstracts addressing any of the questions above or other
related issues. We welcome presentations of completed work or
work-in-progress, as well as papers discussing potential research
directions and surveys of recent developments.
If you would like to present at the workshop, please send an abstract at
most 2 pages long (NIPS Format), excluding citations, PDF preferred, to
[log in to unmask] as soon as possible, and no later than October 31,
2008. Acceptance decisions will be sent on November 7, 2008.
Melissa Carroll, Princeton University
Irina Rish, IBM
Francisco Pereira, Princeton University
Guillermo Cecchi, IBM
Tutorial: Francisco Pereira, Princeton University
Lars Kai Hansen, Technical University of Denmark
Jean-Baptiste Poline/Bertrand Thirion, Neurospin