FMLKNN: A fuzzy membership function based K-Nearest Neighbor approach for multi-label classification

Jung-Yi Jiang, Shie Jue Lee

研究成果: Article同行評審

2 引文 斯高帕斯(Scopus)

摘要

Multi-label classification learning concerns the determination of categories in the situation where one pattern may belong to more than one category. In this paper, we propose a fuzzy membership function based K-Nearest Neighbor algorithm, named FMLKNN, for multi-label classification. For an unseen pattern, the scores of its relevance to the classes are calculated based on a given membership function. The unseen pattern is then decided to belong to every class to which the pattern's membership degree of "belong to the class" is higher than the degree of "not belong to the class". The membership function used can be learned from applying the Multi-Label K-Nearest Neighbor algorithm (MLKNN). Experimental results show that our proposed method achieves better performance than others. ICIC International

原文English
頁(從 - 到)1069-1075
頁數7
期刊ICIC Express Letters
5
發行號4 A
出版狀態Published - 2011 4月 1

All Science Journal Classification (ASJC) codes

  • 控制與系統工程
  • 電腦科學(全部)

指紋

深入研究「FMLKNN: A fuzzy membership function based K-Nearest Neighbor approach for multi-label classification」主題。共同形成了獨特的指紋。

引用此