{"title":"A weighted voting method using minimum square error based on Extreme Learning Machine","authors":"Jingjing Cao, S. Kwong, Ran Wang, Ke Li","doi":"10.1109/ICMLC.2012.6358949","DOIUrl":null,"url":null,"abstract":"Extreme Learning Machine (ELM) has become popular for solving classification problem due to its fast speed. However, the system of ELM may be unreliable since its performance often relies on random input hidden node parameters. The techniques of combining multiple classifiers are widely adopted to improve both reliability and accuracy of a single classifier. Thus, this paper presents a minimum square error (MSE) based weighted voting method to optimize the linear combination of multiple ELMs. The experimental results over ten VCI data sets show better classification performance than the original ELM and the voting based ELM classifiers.","PeriodicalId":128006,"journal":{"name":"2012 International Conference on Machine Learning and Cybernetics","volume":"73 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2012-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"27","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2012 International Conference on Machine Learning and Cybernetics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICMLC.2012.6358949","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 27
Abstract
Extreme Learning Machine (ELM) has become popular for solving classification problem due to its fast speed. However, the system of ELM may be unreliable since its performance often relies on random input hidden node parameters. The techniques of combining multiple classifiers are widely adopted to improve both reliability and accuracy of a single classifier. Thus, this paper presents a minimum square error (MSE) based weighted voting method to optimize the linear combination of multiple ELMs. The experimental results over ten VCI data sets show better classification performance than the original ELM and the voting based ELM classifiers.