DRCW-OVO: Distance-based relative competence weighting combination for One-vs-One strategy in multi-class problems
作者：
Highlights：
• The non-competence is an important problem in One-vs-One strategy.
• We develop a distance-based combination strategy, based on Dynamic Classifier Weighting strategies.
• Weights are settled depending on the closeness of the test instance to each one of the classes
• The effect of the non-competent classifiers is reduced.
• The new strategy enhances the results obtained w.r.t. the state-of-the-art aggregations.
摘要
Highlights•The non-competence is an important problem in One-vs-One strategy.•We develop a distance-based combination strategy, based on Dynamic Classifier Weighting strategies.•Weights are settled depending on the closeness of the test instance to each one of the classes•The effect of the non-competent classifiers is reduced.•The new strategy enhances the results obtained w.r.t. the state-of-the-art aggregations.
论文关键词：Multi-class classification,Pairwise learning,One-vs-One,Decomposition strategies,Ensembles,Classifier selection
论文评审过程：Received 2 October 2013, Revised 30 June 2014, Accepted 18 July 2014, Available online 2 August 2014.
论文官网地址：https://doi.org/10.1016/j.patcog.2014.07.023
