Name of the Event: CALD Seminar
Location of the Event: Newell-Simon Hall  1507

Speaker Name: Tom Dean

Title of Talk: Scalable Inference in Hierarchical Models of the Neocortex

Start Time: 1:30pm


Scheduled Talk:

Note:
Please contact Sharon Cavlovich for appointments via email sharonw@cs.cmu.edu

Abstract:
Borrowing insights from computational neuroscience, we present a class
of generative models well suited to modeling perceptual processes and
an algorithm for learning their parameters that promises to scale to
learning very large models.  The models are hierarchical, composed of
multiple levels, and allow input only at the lowest level, the base of
the hierarchy.  Connections within a level are generally local and may
or may not be directed.  Connections between levels are directed and
generally do not span multiple levels.

The learning algorithm falls within the general family of expectation
maximization algorithms.  Parameter estimation proceeds level-by-level
starting with components in the lowest level and moving up the
hierarchy.  Having learned the parameters for the components in a
given level, those parameters are fixed and needn't be revisited for
the purposes of learning.  These parameters do, however, play an
important role in learning the parameters for higher-level components
by helping to generate the samples used in subsequent parameter
estimation.  Within levels, learning is decomposed into many local
subproblems suggesting a straightforward parallel implementation.

The inference required for learning is carried out by local message
passing and the arrangement of connections within the underlying
networks is designed to facilitate this method of inference.  Learning
is unsupervised but can be easily adapted to accommodate labeled data.
In addition to describing several variants of the basic algorithm, we
present preliminary experimental results demonstrating the
pattern-recognition capabilities of our approach and some of the
characteristics of the approximations that the algorithms produce.



Relevant URL:
http://www.cs.brown.edu/~tld/

Monica Hopes, Executive Assistant
Center for Automated Learning and Discovery (CALD)
Phone (412) 268-5527 Fax: (412) 268-3431 Email: meh@cs.cmu.edu

"Sorrow looks back.  Worry looks around.  Faith looks up."