作者: Yiming Zhou , ChunHui Zhang
DOI: 10.1109/FUZZY.2007.4295569
关键词:
摘要: This paper proposes a weighted fuzzy C-means (W-FPCM) clustering algorithm. It is based on the possibilistic (FPCM) The idea of W-FPCM came from Pareto principle. associates different weights to variables when computing distance in process after filtering out less important variables. algorithm performs well for data sets UCI (University California, Irvine) terms three evaluation methods. first accuracy, second refinement FPCM's objective function; third Kosko's entropy formula. main difference between conventional feature selection algorithms and ours that our weighting scheme runs through while others just