Natural language research announcements

[Rajhans Samdani <rsamdan2 AT illinois.edu>]

Dear Faculty and Students,

This is a gentle reminder about today's AIIS seminar; this week we have a 
talk related to structured prediction by Shankar Vembu who is doing his Post 
Doc in the Cognitive Computation Lab under Prof. Dan Roth. Attached with this 
mail are the title and the abstract of the talk.

Title:
Learning to Predict Combinatorial Structures

Abstract:
The major challenge in designing a discriminative learning
algorithm for predicting structured data is to address the
computational issues arising from the exponential size of the output
space. Existing algorithms make different assumptions to ensure
efficient, polynomial time estimation of model parameters. For several
combinatorial structures, including cycles, partially ordered sets,
permutations and other graph classes, these assumptions do not hold.
In this work, we address the problem of designing learning algorithms
for predicting combinatorial structures by introducing two new
assumptions: (i) The first assumption is that a particular counting
problem can be solved efficiently. The consequence is a generalization
of the classical ridge regression for structured prediction. (ii) The
second assumption is that a particular sampling problem can be solved
efficiently. The consequence is a new technique for designing and
analyzing probabilistic structured prediction models. These results
can be applied to solve several complex learning problems including
but not limited to multi-label classification, multi-category
hierarchical classification, and label ranking.

Hoping to see you all there,
Regards,
Rajhans


Rajhans Samdani,
Graduate Student,
Dept. of Computer Science,
University of Illinois at Urbana-Champaign.