Skip to main navigation Skip to search Skip to main content

Combining complementary neural network and error-correcting output codes for multiclass classification problems

  • Mahidol University
  • Ramkhamhaeng University
  • CHE

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

This paper presented an innovative method, combining Complementary Neural Networks (CMTNN) and Error-Correcting Output Codes (ECOC), to solve multiclass classification problem. CMTNN consist of truth neural network and falsity neural network created based on truth and falsity information, respectively. In the experiment, we deal with feed-forward backpropagation neural networks, trained using 10 fold cross-validation method and classified based on minimum distance. The proposed approach has been tested with three benchmark problems: balance, vehicle and nursery from the UCI machine learning repository. We found that our approach provides better performance compared to the existing techniques considering on either CMTNN or ECOC.

Original languageEnglish
Title of host publication10th WSEAS International Conference on Applied Computer and Applied Computational Science, ACACOS'11
Pages49-54
Number of pages6
Publication statusPublished - 2011
Externally publishedYes
Event10th WSEAS International Conference on Applied Computer and Applied Computational Science, ACACOS'11 - Venice, Italy
Duration: 8 Mar 201110 Mar 2011

Publication series

Name10th WSEAS International Conference on Applied Computer and Applied Computational Science, ACACOS'11

Conference

Conference10th WSEAS International Conference on Applied Computer and Applied Computational Science, ACACOS'11
Country/TerritoryItaly
CityVenice
Period8/03/1110/03/11

Keywords

  • Complementary neural network
  • Error-Correcting Output Codes (ECOC)
  • Feed-forward backpropagation neural network
  • Multicass classification

Fingerprint

Dive into the research topics of 'Combining complementary neural network and error-correcting output codes for multiclass classification problems'. Together they form a unique fingerprint.

Cite this