Revisiting output coding for sequential supervised learning Public Deposited

http://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/fx719p77g

Descriptions

Attribute NameValues
Creator
Abstract or Summary
  • Markov models are commonly used for joint inference of label sequences. Unfortunately, inference scales quadratically in the number of labels, which is problematic for training methods where inference is repeatedly preformed and is the primary computational bottleneck for large label sets. Recent work has used output coding to address this issue by converting a problem with many labels to a set of problems with binary labels. Models were independently trained for each binary problem, at a much reduced computational cost, and then combined for joint inference over the original labels. Here we revisit this idea and show through experiments on synthetic and benchmark data sets that the approach can perform poorly when it is critical to explicitly capture the Markovian transition structure of the large-label problem. We then describe a simple cascade-training approach and show that it can improve performance on such problems with negligible computational overhead.
Resource Type
Date Available
Date Copyright
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Academic Affiliation
Non-Academic Affiliation
Keyword
Subject
Rights Statement
Language
Replaces
Additional Information
  • description.provenance : Approved for entry into archive by Julie Kurtz(julie.kurtz@oregonstate.edu) on 2009-03-06T00:22:50Z (GMT) No. of bitstreams: 1 Hao_thesis.pdf: 350533 bytes, checksum: 4f688db800bd1715370e8e0b5ac09224 (MD5)
  • description.provenance : Submitted by Guohua Hao (haog@onid.orst.edu) on 2009-03-03T20:38:28Z No. of bitstreams: 1 Hao_thesis.pdf: 350533 bytes, checksum: 4f688db800bd1715370e8e0b5ac09224 (MD5)
  • description.provenance : Made available in DSpace on 2009-03-10T22:48:17Z (GMT). No. of bitstreams: 1 Hao_thesis.pdf: 350533 bytes, checksum: 4f688db800bd1715370e8e0b5ac09224 (MD5)
  • description.provenance : Approved for entry into archive by Laura Wilson(laura.wilson@oregonstate.edu) on 2009-03-10T22:48:17Z (GMT) No. of bitstreams: 1 Hao_thesis.pdf: 350533 bytes, checksum: 4f688db800bd1715370e8e0b5ac09224 (MD5)

Relationships

Parents:

This work has no parents.

Last modified

Downloadable Content

Download PDF

Items