Search |  Contact |  SRI Home Do not follow this link, or your host will be blocked from this site. This is a spider trap. Do not follow this link, or your host will be blocked from this site. This is a spider trap. Do not follow this link, or your host will be blocked from this site. This is a spider trap.A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A ASRI International.  333 Ravenswood Avenue.  Menlo Park, CA 94025-3493. SRI International is a nonprofit corporation.

Publication in BibTeX Format

@ARTICLE{AICPub1807:2010, AUTHOR={Madani, O. and Huang, J.}, TITLE={Large-Scale Many-Class Prediction via Flat Techniques}, JOURNAL={Workshop on Large-Scale Hierarchical Text Classification at ECIR}, YEAR={2010}, KEYWORDS={large-scale learning, many-class learning, multiclass learning, text classification, convergence, mistake-bounds}, ABSTRACT={Prediction problems with huge numbers of classes are becoming more common. While class taxonomies are available in certain cases, we have observed that simple flat learning and classification, via index learning and related techniques, offers significant efficiency and accuracy advantages. In the PASCAL challenge on large-scale hierarchical text classification, the accuracies we obtained ranked in the top three in all evaluations. We also found that using committees of a few learned models boosted accuracy, and observed a tradeoff between accuracy versus memory and time efficiency. This paper is an extension of our short report paper on the competition, and includes a proof of convergence, with a mistake bound, for a variant of indexing in the two-class setting for the separable case.} }

SRI International
©2014 SRI International 333 Ravenswood Avenue, Menlo Park, CA 94025-3493
SRI International is an independent, nonprofit corporation. Privacy policy