Supervised fuzzy partitioning
作者：
Highlights：
• A framework is presented, extending k-means clustering to be applicable to supervised tasks.
• We adopt an entropy approach to fuzzification and feature weighting.
• An efficient block coordinate descent scheme is formulated to find local minima.
• A flexible, nonlinear classifier is presented, capable of handling high-dimensional settings.
• Experimental results show the superior performance of the proposed method over state-of-the-art classifiers.
摘要
•A framework is presented, extending k-means clustering to be applicable to supervised tasks.•We adopt an entropy approach to fuzzification and feature weighting.•An efficient block coordinate descent scheme is formulated to find local minima.•A flexible, nonlinear classifier is presented, capable of handling high-dimensional settings.•Experimental results show the superior performance of the proposed method over state-of-the-art classifiers.
论文关键词：Supervised k-means,Centroid-based clustering,Entropy-based regularization,Feature weighting,Mixtures of experts
论文评审过程：Received 21 October 2018, Revised 11 May 2019, Accepted 17 August 2019, Available online 19 August 2019, Version of Record 29 August 2019.
论文官网地址：https://doi.org/10.1016/j.patcog.2019.107013
