[Colloquium] TTIC Colloquium: Prof. Richard Zemel (Univ of Toronto)

[Dawn Ellis]

When:     Tuesday, June 4th at 11 a.m.

Where:    TTIC, 6045 S Kenwood Avenue, 5th Floor, Room #526

Who:       Prof. Richard Zemel, University of Toronto

Title:        Integrating Learning in Structured Prediction and Latent
               Variable Models

Abstract:

The most powerful machine learning methods involve learning high-dimensional
and often complicated latent variables. Learning in these models, such as
neural networks and Boltzmann machines, typically operates with little
structure assumed about the latent variables. Structured prediction
problems, on the other hand, typically make very limited use of latent
variables. I will discuss how both of these types of models can operate
synergistically.  I will show how latent-variable models can be integrated
into structured prediction problems, and will also illustrate useful
applications of structured prediction methods in learning latent-variable
models.

Host: Tamir Hazan, tamir at ttic.edu

-- 
*Dawn Ellis*
Administrative Assistant
773-834-1757
dellis at ttic.edu

TTIC
6045 S. Kenwood Ave.
Chicago, IL. 60637
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20130529/a418cc26/attachment.htm