Connectionists: NIPS satellite meeting on deep learning

[Yoshua Bengio]

Hello,

Following up on Geoffrey Hinton's Monday NIPS tutorial on Deep Belief  
Nets, there will be a satellite meeting/workshop on Thursday  
afternoon (i.e. not at the usual time for workshops) at the Vancouver  
Hyatt. This Deep Learning workshop
        http://www.iro.umontreal.ca/~lisa/deepNIPS2007
is part of the "Neuro-Thursday" this year
       http://nips.cc/Conferences/2007/Neuro-Thursday
SEATING IS LIMITED at the Hyatt for this workshop, so if you are  
interested (check the program at the URL above) you are encouraged to  
register as soon as possible.

For interested participants of the deep learning workshop buses to  
Whistler are scheduled after the workshop (the regular buses to  
Whisler are at 2pm). The workshop also honors Geoffrey Hinton's 60-th  
birthday.

More on the meeting:

Theoretical results strongly suggest that in order to learn the kind  
of complicated functions that can represent high-level abstractions  
(e.g. in vision, language, and other AI-level tasks), one may need  
"deep architectures", which are composed of multiple levels of non- 
linear operations (such as in neural nets with many hidden layers).  
Searching the parameter space of deep architectures is a difficult  
optimization task, but learning algorithms (e.g. Deep Belief  
Networks) have recently been proposed to tackle this problem with  
notable success, beating the state-of-the-art in certain areas.
This workshop is intended to bring together researchers interested in  
the question of deep learning in order to review the current  
algorithms' principles and successes, but also to identify the  
challenges, and to formulate promising directions of investigation.  
Besides the algorithms themselves, there are many fundamental  
questions that need to be addressed: What would be a good  
formalization of deep learning? What new ideas could be exploited to  
make further inroads to that difficult optimization problem? What  
makes a good high-level representation or abstraction? What type of  
problem is deep learning appropriate for?

Schedule:

2:00pm - 2:25pm Yee-Whye Teh, Gatsby Unit : Deep Belief Networks
2:25pm - 2:45pm John Langford, Yahoo Research: Theoretical Results on  
Deep Architectures
2:45pm - 3:05pm Yoshua Bengio, University of Montreal: Optimizing  
Deep Architectures
3:05pm - 3:25pm Yann Le Cun, New York University: Learning deep  
hierarchies of invariant features
3:25pm - 3:45pm Martin Szummer, Microsoft Research: Deep networks for  
information retrieval
3:45pm - 4:00pm Coffee break
4:00pm - 4:20pm Max Welling, University of California: Hierarchical  
Representations from networks of HDPs
4:20pm - 4:40pm Andrew Ng, Stanford University: Self-taught learning:  
Transfer learning from unlabeled data
4:40pm - 5:10pm Geoff Hinton, University of Toronto: Restricted  
Boltzmann machines with multiplicative interactions
5:10pm - 5:30pm Discussion

-- Yoshua Bengio