聚類中心的初始化方法
A NEW INITIALIZATION METHOD OF CLUSTER CENTERS
-
摘要: 本文對(duì)用于聚類中心初始化的勢(shì)函數(shù)的幾個(gè)參數(shù)選擇問(wèn)題進(jìn)行了討論,給出了這些參數(shù)的兩種形式。同時(shí)提出了一種新的使用密度函數(shù)法進(jìn)行聚類中心初始化的方法,進(jìn)行了一組對(duì)比實(shí)驗(yàn),得到了令人滿意的結(jié)果。Abstract: The problems of parameter selections for potential function used to initialize cluster centers are discussed, and two methods are given for determining these parameters. Then a new density function to initialize cluster centers is also given which is computational effective. Finally, a set of compared experiments is presented to show the effectiveness of the proposed methods.
-
Bezdek J C. Pattern Recognition with Fuzzy Objective Function Algorithms. New York: Plenum, 1981, 43-93.[2]Bezdek J C, Hathaway R, Sabin M, Tucker W. Convergence theory for fuzzy C-means: Coun-[3]terexample and repairs. IEEE Trans. on SMC., 1987, SMC-17(5): 873-877.[4]Yager R R, Filev D P. Approximate clustering via the mountain method. IEEE Trans. on SMC., 1994, SMC-24(8): 1279-1284.[5]Chiu S L. Fuzzy model identification based on cluster estimation. Journal of Intelligent and Fuzzy Systems, 1994, 2(3): 267-278.[6]Moser C A, Scott W. British Towns. Edinburgh: Oliver and Boyd, 1961.[7]裴繼紅.基于模糊信息處理的圖像分割方法研究:[博士論文].西安電子科技大學(xué),1998. -
計(jì)量
- 文章訪問(wèn)數(shù): 2918
- HTML全文瀏覽量: 153
- PDF下載量: 541
- 被引次數(shù): 0
下載:
