Connectionists: Non-linear dimensionality reduction using neural networks

[Ruslan Salakhutdinov]

Dear Colleagues,

I wanted to advertise our paper appearing in this issue of Science
magazine (July 28, 2006). The paper, along with supporting material and
Matlab code, is publicly available at:
http://www.cs.toronto.edu/~rsalakhu/index.html
or from Science online (subscription-only) at
http://www.sciencemag.org/cgi/content/full/313/5786/504


Here is the abstract;

Reducing the Dimensionality of Data with Neural Networks
G. E. Hinton and R. R. Salakhutdinov

High-dimensional data can be converted to low-dimensional codes by
training a multilayer neural network with a small central layer to
reconstruct high-dimensional input vectors. Gradient descent can be used
for fine-tuning the weights in such "autoencoder" networks, but this works
well only if the initial weights are close to a good solution. We describe
an effective way of initializing the weights that allows deep autoencoder
networks to learn low-dimensional codes that work much better than
principal components analysis as a tool to reduce the dimensionality of
data.

Department of Computer Science, University of Toronto, 6 King's College
Road, Toronto, Ontario M5S 3G4, Canada.

Science 28 July 2006:
Vol. 313. no. 5786, pp. 504 - 507


Sincerely,
Russ