A new dynamic ensemble selection method for numeral recognition

被引：0
|
作者：
Ko, Albert Hung-Ren [1 ]
Sabourin, Robert [1 ]
de Scuza Britto, Alceu, Jr. [2 ]
机构：
[1] Univ Quebec, LIVIA, ETS, 1100 Notre-Dame W St, Montreal, PQ H3C 1K3, Canada
[2] Pontif Cathol Univ, PPGIA, Curitiba, Parana, Brazil
来源：
MULTIPLE CLASSIFIER SYSTEMS, PROCEEDINGS | 2007年 / 4472卷
基金：
加拿大自然科学与工程研究理事会;
关键词：
fusion function; combining classifiers; diversity confusion matrix; pattern recognition; majority voting; ensemble of learning machines;
D O I：
暂无
中图分类号：
TP18 [人工智能理论];
学科分类号：
081104 ; 0812 ; 0835 ; 1405 ;
摘要：
An ensemble of classifiers (EoC) has been shown to he effective in improving classifier performance. To optimize EoC, the ensemble selection is one of the most imporatant issues. Dynamic scheme urges the use of different ensembles for different samples, but it has been shown that dynamic selection does not give better performance than static selection. We propose a dynamic selection scheme which explores the property of the oracle concept. The result suggests that the proposed scheme is apparently better than the selection based on popular majority voting error.
引用
收藏
页码：431 / +
页数：2
相关论文
共 50 条
  • [1] A new HMM-based ensemble generation method for numeral recognition
    Ko, Albert Hung-Ren
    Sabourin, Robert
    de Souza Britto, Alceu
    [J]. MULTIPLE CLASSIFIER SYSTEMS, PROCEEDINGS, 2007, 4472 : 52 - +
  • [2] New Structural Method for Numeral Recognition
    Benchaou, Soukaina
    Nasri, M'Barek
    El Melhaoui, Ouafae
    [J]. PROCEEDINGS OF THE MEDITERRANEAN CONFERENCE ON INFORMATION & COMMUNICATION TECHNOLOGIES 2015, VOL 1, 2016, 380 : 237 - 245
  • [3] An Experimental evaluation of Feature selection based Classifier Ensemble for Handwritten Numeral Recognition
    Singh, Pratibha
    Verma, Ajay
    Chaudhari, Narendra S.
    [J]. 2014 INTERNATIONAL CONFERENCE ON ELECTRONICS AND COMMUNICATION SYSTEMS (ICECS), 2014,
  • [4] A dynamic selection ensemble method for target recognition based on clustering and randomized reference classifier
    Xueman Fan
    Shengliang Hu
    Jingbo He
    [J]. International Journal of Machine Learning and Cybernetics, 2019, 10 : 515 - 525
  • [5] A dynamic selection ensemble method for target recognition based on clustering and randomized reference classifier
    Fan, Xueman
    Hu, Shengliang
    He, Jingbo
    [J]. INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS, 2019, 10 (03) : 515 - 525
  • [6] Emotion Recognition Based on Dynamic Ensemble Feature Selection
    Yang, Yong
    Wang, Guoyin
    Kong, Hao
    [J]. MAN-MACHINE INTERACTIONS, 2009, 59 : 217 - 225
  • [7] Study on Handwritten Numeral Recognition Method
    Wang, Yan
    [J]. 2015 SSR International Conference on Social Sciences and Information (SSR-SSI 2015), Pt 2, 2015, 11 : 784 - 787
  • [8] New Dynamic Ensemble of Classifiers Selection Approach Based on Confusion matrix for Arabic Handwritten recognition
    Nabiha, Azizi
    Nadir, Farah
    [J]. 2012 INTERNATIONAL CONFERENCE ON MULTIMEDIA COMPUTING AND SYSTEMS (ICMCS), 2012, : 308 - 313
  • [9] A new method for weighted ensemble clustering and coupled ensemble selection
    Banerjee, Arko
    Pujari, Arun K.
    Panigrahi, Chhabi Rani
    Pati, Bibudhendu
    Nayak, Suvendu Chandan
    Weng, Tien-Hsiung
    [J]. CONNECTION SCIENCE, 2021, 33 (03) : 623 - 644
  • [10] A New Way for Handwritten Numeral Recognition
    Guan, Baolin
    Ba, Lideng
    [J]. ADVANCES IN APPLIED SCIENCE AND INDUSTRIAL TECHNOLOGY, PTS 1 AND 2, 2013, 798-799 : 643 - 646
12345