计算机应用 ›› 2011, Vol. 31 ›› Issue (09): 2538-2541.DOI: 10.3724/SP.J.1087.2011.02538

• 人工智能 • 上一篇    下一篇

基于核的直觉模糊聚类算法

范成礼,雷英杰   

  1. 空军工程大学 导弹学院, 陕西 三原 713800
  • 收稿日期:2011-03-28 修回日期:2011-05-03 发布日期:2011-09-01 出版日期:2011-09-01
  • 通讯作者: 范成礼
  • 作者简介:范成礼(1988-),女,四川眉山人,硕士研究生,主要研究方向:智能信息处理、信息融合;
    雷英杰(1956-),男,陕西渭南人,教授,博士,博士生导师,主要研究方向:智能信息处理、智能决策。
  • 基金资助:
    国家自然科学基金资助项目(60773209)

Kernel based intuitionistic fuzzy clustering algorithm

FAN Cheng-li,LEI Ying-jie   

  1. Missile Institute, Air Force Engineering University, Sanyuan Shaanxi 713800, China
  • Received:2011-03-28 Revised:2011-05-03 Online:2011-09-01 Published:2011-09-01
  • Contact: FAN Cheng-li

摘要: 针对现有的直觉模糊聚类算法性能的问题,提出一种基于核的直觉模糊聚类算法(IFKCM)。该算法引入高斯核函数,将直觉模糊集合从原始观察空间映射到高维特征空间,减少了计算时间且提高了聚类精度;同时改进了现有的直觉模糊聚类算法中的概率型约束条件,使其对噪声和野值点具有较好的鲁棒性。最后,通过实际数据和人工数据与常用聚类算法进行了对比实验,结果表明该算法较大幅度地提高了直觉模糊聚类算法的性能。

关键词: 直觉模糊集, 直觉模糊聚类, 模糊核C-均值, 核函数, 高斯核函数

Abstract: A kernel based intuitionistic fuzzy clustering algorithm named IFKCM was proposed on the basis of analyzing the deficiency of the existing clustering algorithm. The new algorithm, through introducing Gauss kernel, mapped the intuitionistic fuzzy sets from their original space to a high dimensional space (or kernel space), so as to have shorter computational time and more accurate result. Besides, it was robust to the noises because it improved the constraint conditions used in the existing intuitionistic fuzzy clustering algorithm. Finally, compared with the traditional algorithm, the proposed algorithm has made some significant progress, and the experimental result has proved its effectiveness.

Key words: intuitionistic fuzzy set, intuitionistic fuzzy clustering, Fuzzy Kernel C-Means (FKCM), kernel function, Gauss kernel function

中图分类号: