Parameter-free classification in multi-class imbalanced data sets

被引:20
作者
Cerf, Loic [1 ]
Gay, Dominique [2 ]
Selmaoui-Folcher, Nazha [3 ]
Cremilleux, Bruno [4 ]
Boulicaut, Jean-Francois [5 ]
机构
[1] Univ Fed Minas Gerais, Dept Comp Sci, Belo Horizonte, MG, Brazil
[2] Orange Labs, F-22307 Lannion, France
[3] Univ New Caledonia, PPME EA3325, Noumea, New Caledonia
[4] Univ Caen, GREYC CNRS UMR6072, F-14032 Caen, France
[5] Univ Lyon, CNRS, INRIA, INSA Lyon,LIRIS,UMR5205, F-69621 Villeurbanne, France
关键词
Classification; Association rules; Multi-class context; Imbalanced data set; One-Versus-Each framework; DISCOVERY; PATTERNS; SMOTE;
D O I
10.1016/j.datak.2013.06.001
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Many applications deal with classification in multi-class imbalanced contexts. In such difficult situations, classical CBA-like approaches (Classification Based on Association rules) show their limits. Most CBA-like methods actually are One-Vs-All approaches (OVA), i.e., the selected classification rules are relevant for one class and irrelevant for the union of the other classes. In this paper, we point out recurrent problems encountered by OVA approaches applied to multi-class imbalanced data sets (e.g., improper bias towards majority classes, conflicting rules). That is why we propose a new One-Versus-Each (OVE) framework. In this framework, a rule has to be relevant for one class and irrelevant for every other class taken separately. Our approach, called fitcare, is empirically validated on various benchmark data sets and our theoretical findings are confirmed. (C) 2013 Elsevier B.V. All rights reserved.
引用
收藏
页码:109 / 129
页数:21
相关论文
共 50 条
  • [31] A Novel and Effective Multi-Class Classification Method for Imbalanced Medical Transcriptions
    Bhardwaj, Priti
    Baliyan, Niyati
    IETE JOURNAL OF RESEARCH, 2024, 8 (6734-6744) : 6734 - 6744
  • [32] SAMME.C2 algorithm for imbalanced multi-class classification
    So, Banghee
    Valdez, Emiliano A.
    Soft Computing, 2024, 28 (17-18) : 9387 - 9404
  • [33] PF-SMOTE: A novel parameter-free SMOTE for imbalanced datasets
    Chen, Qiong
    Zhang, Zhong-Liang
    Huang, Wen-Po
    Wu, Jian
    Luo, Xing-Gang
    NEUROCOMPUTING, 2022, 498 : 75 - 88
  • [34] A Hybrid Sampling Approach for Imbalanced Binary and Multi-Class Data Using Clustering Analysis
    Palli, Abdul Sattar
    Jaafar, Jafreezal
    Hashmani, Manzoor Ahmed
    Gomes, Heitor Murilo
    Gilal, Abdul Rehman
    IEEE ACCESS, 2022, 10 : 118639 - 118653
  • [35] Online active learning method for multi-class imbalanced data stream
    Li, Ang
    Han, Meng
    Mu, Dongliang
    Gao, Zhihui
    Liu, Shujuan
    KNOWLEDGE AND INFORMATION SYSTEMS, 2024, 66 (04) : 2355 - 2391
  • [36] CLASSIFICATION OF LIDAR DATA BASED ON MULTI-CLASS SVM
    Samadzadegan, F.
    Bigdeli, B.
    Ramzi, P.
    2010 CANADIAN GEOMATICS CONFERENCE AND SYMPOSIUM OF COMMISSION I, ISPRS CONVERGENCE IN GEOMATICS - SHAPING CANADA'S COMPETITIVE LANDSCAPE, 2010, 38
  • [37] Feature selection and its combination with data over-sampling for multi-class imbalanced datasets
    Tsai, Chih-Fong
    Chen, Kuan-Chen
    Lin, Wei -Chao
    APPLIED SOFT COMPUTING, 2024, 153
  • [38] Evolutionary inversion of class distribution in overlapping areas for multi-class imbalanced learning
    Fernandes, Everlandio R. Q.
    de Carvalho, Andre C. P. L. F.
    INFORMATION SCIENCES, 2019, 494 : 141 - 154
  • [39] Accurate and efficient sequential ensemble learning for highly imbalanced multi-class data
    Vong, Chi-Man
    Du, Jie
    NEURAL NETWORKS, 2020, 128 : 268 - 278
  • [40] Study of Multi-Class Classification Algorithms' Performance on Highly Imbalanced Network Intrusion Datasets
    Bulavas, Viktoras
    Marcinkevicius, Virginijus
    Ruminski, Jacek
    INFORMATICA, 2021, 32 (03) : 441 - 475