Abstract
We describe and analyze an algorithmic framework for online classification where each online trial consists of multiple prediction tasks that are tied together. We tackle the problem of updating the online hypothesis by defining a projection problem in which each prediction task corresponds to a single linear constraint. These constraints are tied together through a single slack parameter. We then introduce a general method for approximately solving the problem by projecting simultaneously and independently on each constraint which corresponds to a prediction sub-problem, and then averaging the individual solutions. We show that this approach constitutes a feasible, albeit not necessarily optimal, solution for the original projection problem. We derive concrete simultaneous projection schemes and analyze them in the mistake bound model. We demonstrate the power of the proposed algorithm in experiments with online multiclass text categorization. Our experiments indicate that a combination of class-dependent features with the simultaneous projection method outperforms previously studied algorithms.
Original language | English |
---|---|
Title of host publication | NIPS 2006 |
Subtitle of host publication | Proceedings of the 19th International Conference on Neural Information Processing Systems |
Editors | Bernhard Scholkopf, John C. Platt, Thomas Hofmann |
Publisher | MIT Press Journals |
Pages | 17-24 |
Number of pages | 8 |
ISBN (Electronic) | 0262195682, 9780262195683 |
State | Published - 2006 |
Event | 19th International Conference on Neural Information Processing Systems, NIPS 2006 - Vancouver, Canada Duration: 4 Dec 2006 → 7 Dec 2006 |
Publication series
Name | NIPS 2006: Proceedings of the 19th International Conference on Neural Information Processing Systems |
---|
Conference
Conference | 19th International Conference on Neural Information Processing Systems, NIPS 2006 |
---|---|
Country/Territory | Canada |
City | Vancouver |
Period | 4/12/06 → 7/12/06 |
Bibliographical note
Publisher Copyright:© NIPS 2006.All rights reserved