【摘要】 首先对模糊C-均值聚类算法做了简要分析和评论,根据其特点,提出了一种基于遗传模拟退火算法的聚类分析方法,从而提高了遗传算法的全局搜索能力。算法中采用了适合于模糊聚类的树型编码方案,实验表明,该算法可克服系统对数据集及初始聚类中心的敏感性,避免陷入局部极小,具有良好的准确性与可靠性,在模式识别、数据挖掘等领域有着广泛的应用前景。更多还原

【Abstract】 The Fuzzy C- means clustering algothrim was described and remarked, then,according to its characteristic ,an improvedclusting analysis method based on genetic simulated annealing algothrim is proposed, thus raised overall search ability of genetic al-gorithm。A tree encoding scheme which is proper for the fuzzy clustering algothrims is determined,Experiment results show that theimproved algothrim give us more satisfactory results by escaping from local minima regardless of the data set and the selection of theinitial clustering centers, and have good accuracy and dependability。 It can be used in many fields such as pattern recognition anddata mining。更多还原