Improving kNN multi-label classification in Prototype Selection scenarios using class proposals
作者：
Highlights：
• Improving Prototype Selection-based classification proposing likely labels of the reduced set.
• kNN search within the original training set restricted to those proposed labels.
• Scheme that provides a broad range of solution in the trade-off accuracy efficiency.
• Cost reduction in multi-label classification scenarios and robustness against noise.
• Our approach gets to reach accuracy of kNN with barely a third of distances computed.
摘要
Highlights•Improving Prototype Selection-based classification proposing likely labels of the reduced set.•kNN search within the original training set restricted to those proposed labels.•Scheme that provides a broad range of solution in the trade-off accuracy efficiency.•Cost reduction in multi-label classification scenarios and robustness against noise.•Our approach gets to reach accuracy of kNN with barely a third of distances computed.
论文关键词：K-Nearest Neighbor,Multi-label classification,Prototype Selection,Class proposals
论文评审过程：Received 18 July 2014, Revised 18 November 2014, Accepted 22 November 2014, Available online 3 December 2014.
论文官网地址：https://doi.org/10.1016/j.patcog.2014.11.015
