Prototype Selection for Multilabel Instance-Based Learning

被引:3
|
作者
Filippakis, Panagiotis [1 ]
Ougiaroglou, Stefanos [1 ]
Evangelidis, Georgios [2 ]
机构
[1] Int Hellen Univ, Dept Informat & Elect Engn, Sch Engn, Thessaloniki 57400, Greece
[2] Univ Macedonia, Sch Informat Sci, Dept Appl Informat, 156 Egnatia St, Thessaloniki 54636, Greece
关键词
data reduction techniques; instance reduction; multilabel classification; prototype selection; instance-based classification; binary relevance; CNN; IB2; BRkNN; DATA REDUCTION; LOCAL SETS; CLASSIFICATION; GENERATION; ALGORITHM; KNN;
D O I
10.3390/info14100572
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Reducing the size of the training set, which involves replacing it with a condensed set, is a widely adopted practice to enhance the efficiency of instance-based classifiers while trying to maintain high classification accuracy. This objective can be achieved through the use of data reduction techniques, also known as prototype selection or generation algorithms. Although there are numerous algorithms available in the literature that effectively address single-label classification problems, most of them are not applicable to multilabel data, where an instance can belong to multiple classes. Well-known transformation methods cannot be combined with a data reduction technique due to different reasons. The Condensed Nearest Neighbor rule is a popular parameter-free single-label prototype selection algorithm. The IB2 algorithm is the one-pass variation of the Condensed Nearest Neighbor rule. This paper proposes variations of these algorithms for multilabel data. Through an experimental study conducted on nine distinct datasets as well as statistical tests, we demonstrate that the eight proposed approaches (four for each algorithm) offer significant reduction rates without compromising the classification accuracy.
引用
收藏
页数:22
相关论文
共 50 条
  • [1] Combining instance-based learning and logistic regression for multilabel classification
    Cheng, Weiwei
    Huellermeier, Eyke
    MACHINE LEARNING, 2009, 76 (2-3) : 211 - 225
  • [2] Combining Instance-Based Learning and Logistic Regression for Multilabel Classification
    Cheng, Weiwei
    Huellermeier, Eyke
    MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES, PT I, 2009, 5781 : 6 - 6
  • [3] Combining instance-based learning and logistic regression for multilabel classification
    Weiwei Cheng
    Eyke Hüllermeier
    Machine Learning, 2009, 76 : 211 - 225
  • [4] Advances in Instance Selection for Instance-Based Learning Algorithms
    Henry Brighton
    Chris Mellish
    Data Mining and Knowledge Discovery, 2002, 6 : 153 - 172
  • [5] Advances in instance selection for instance-based learning algorithms
    Brighton, H
    Mellish, C
    DATA MINING AND KNOWLEDGE DISCOVERY, 2002, 6 (02) : 153 - 172
  • [6] A cooperative coevolutionary algorithm for instance selection for instance-based learning
    Nicolás García-Pedrajas
    Juan Antonio Romero del Castillo
    Domingo Ortiz-Boyer
    Machine Learning, 2010, 78 : 381 - 420
  • [7] A cooperative coevolutionary algorithm for instance selection for instance-based learning
    Garcia-Pedrajas, Nicolas
    Antonio Romero del Castillo, Juan
    Ortiz-Boyer, Domingo
    MACHINE LEARNING, 2010, 78 (03) : 381 - 420
  • [8] Combining three strategies for evolutionary instance selection for instance-based learning
    de Haro-Garcia, Aida
    Perez-Rodriguez, Javier
    Garcia-Pedrajas, Nicolas
    SWARM AND EVOLUTIONARY COMPUTATION, 2018, 42 : 160 - 172
  • [9] Instance selection based on boosting for instance-based learners
    de Haro-Garcia, Aida
    Cerruela-Garcia, Gonzalo
    Garcia-Pedrajas, Nicolas
    PATTERN RECOGNITION, 2019, 96
  • [10] Making Instance-based Learning Theory usable and understandable: The Instance-based Learning Tool
    Dutt, Varun
    Gonzalez, Cleotilde
    COMPUTERS IN HUMAN BEHAVIOR, 2012, 28 (04) : 1227 - 1240