ePrints@IIScePrints@IISc Home | About | Browse | Latest Additions | Advanced Search | Contact | Help

A sequential dual method for structural SVMs

Balamurugan, P and Shevade, Shirish and Sundararajan, S and Keerthi, Sathiya S (2011) A sequential dual method for structural SVMs. In: 2011 SIAM International Conference on Data Mining (SDM), April 28-30,2011, Mesa, Arizona, USA.

[img] PDF
SIAM_Int_Con_Dat_Min_223_2011.pdf - Published Version
Restricted to Registered users only

Download (340kB) | Request a copy
Official URL: http://www.siam.org/meetings/sdm11/


In many real world prediction problems the output is a structured object like a sequence or a tree or a graph. Such problems range from natural language processing to compu- tational biology or computer vision and have been tackled using algorithms, referred to as structured output learning algorithms. We consider the problem of structured classifi- cation. In the last few years, large margin classifiers like sup-port vector machines (SVMs) have shown much promise for structured output learning. The related optimization prob -lem is a convex quadratic program (QP) with a large num-ber of constraints, which makes the problem intractable for large data sets. This paper proposes a fast sequential dual method (SDM) for structural SVMs. The method makes re-peated passes over the training set and optimizes the dual variables associated with one example at a time. The use of additional heuristics makes the proposed method more efficient. We present an extensive empirical evaluation of the proposed method on several sequence learning problems.Our experiments on large data sets demonstrate that the proposed method is an order of magnitude faster than state of the art methods like cutting-plane method and stochastic gradient descent method (SGD). Further, SDM reaches steady state generalization performance faster than the SGD method. The proposed SDM is thus a useful alternative for large scale structured output learning.

Item Type: Conference Paper
Publisher: Society for Industrial and Applied Mathematics
Additional Information: Copyright of this article belongs to Society for Industrial and Applied Mathematics.
Department/Centre: Division of Electrical Sciences > Computer Science & Automation
Date Deposited: 19 Mar 2013 05:57
Last Modified: 19 Mar 2013 05:57
URI: http://eprints.iisc.ac.in/id/eprint/46032

Actions (login required)

View Item View Item