Relaxation of hard classification targets for LSE minimization

Kar Ann Toh, Xudong Jiang, Wei Yun Yau

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In the spirit of stabilizing a solution to handle possible over-fitting of data which is especially common for high order models, we propose a relaxed target training method for regression models which are linear in parameters. This relaxation of training target from the conventional binary values to disjoint classification spaces provides good classification fidelity according to a threshold treatment during the decision process. A particular design to relax the training target is provided under practical consideration. Extension to multiple class problems is formulated before the method is applied to a plug-in full multivariate polynomial model and a reduced model on synthetic data sets to illustrate the idea. Additional experiments were performed using real-world data from the UCI[1] data repository to derive certain empirical evidence.

Original languageEnglish
Title of host publicationEnergy Minimization Methods in Computer Vision and Pattern Recognition - 5th International Workshop, EMMCVPR 2005, Proceedings
Pages187-202
Number of pages16
DOIs
Publication statusPublished - 2005
Event5th International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition, EMMCVPR 2005 - St. Augustine, FL, United States
Duration: 2005 Nov 92005 Nov 11

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3757 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other5th International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition, EMMCVPR 2005
CountryUnited States
CitySt. Augustine, FL
Period05/11/905/11/11

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Relaxation of hard classification targets for LSE minimization'. Together they form a unique fingerprint.

  • Cite this

    Toh, K. A., Jiang, X., & Yau, W. Y. (2005). Relaxation of hard classification targets for LSE minimization. In Energy Minimization Methods in Computer Vision and Pattern Recognition - 5th International Workshop, EMMCVPR 2005, Proceedings (pp. 187-202). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3757 LNCS). https://doi.org/10.1007/11585978_13