Connectionists: CFP: NIPS 2005 Workshop: Large Scale Kernel Machines

[Leon Bottou]

###########################################################

               NIPS 2005 Workshop

            LARGE SCALE KERNEL MACHINES
         <http://nipsworkshop.bottou.org/>

###########################################################

Datasets with millions of observations can be gathered by 
crawling the web, mining business databases, or connecting a 
cheap video tuner to a laptop. Vastly more ambitious learning 
systems are theoretically possible. The literature shows no 
shortage of ideas for sophisticated statistical models. The 
computational cost of learning algorithms is now the bottleneck. 
During the last decade, dataset size has outgrown processor speed. 
Meanwhile, machine learning algorithms became more principled, 
and also more computationally expensive.
 
The workshop investigates computationally efficient ways to 
exploit such large datasets using kernel machines. It will 
show how adequately designed kernel machines can efficiently 
process millions of examples. It will also debate whether 
kernel machines are the best way to achieve such objectives.
 
TOPICS:

* Fast implementation of ordinary Support Vector Machines. 
  How to improve the optimization algorithms and to 
  distribute them on several computers?
 
* Kernel algorithms specifically designed for large scale datasets. 
  For instance, online kernel algorithms are less hungry for memory. 
  Does this improvement comes for free or does it increases the error rates?
 
* Methods for containing the growth of the number of support vectors. 
  Does the number of Support Vectors always grow linearly with the number 
  of examples, as in ordinary Support Vector Machines?

* Comparing the relative strengths of kernel and non kernel methods  
  on large scale datasets. Are kernel methods the best tools 
  for such datasets?

CALL FOR PARTICIPATION:

If you wish to make a presentation,
send a plain text email to <nipsworkshop at bottou.org> 
with title, authors, and a brief abstract (less than one page.)
Please send us this information before November 1st.

ORGANIZERS:

Leon Bottou (NEC Labs, Princeton) 
Olivier Chapelle (MPI, Tuebingen)
Dennis Decoste (Yahoo!, Sunnyvale)
Jason Weston (NEC Labs, Princeton) 



-------------------------------------------------------