论文部分内容阅读
分类属性数据的样本间的分布不平衡、样本的分布与空间距离无关的特点与量子力学中粒子的分布状态由能量决定、粒子分布具有不平衡性的特点相似。基于此,参照量子聚类QC算法确定聚类中心的聚类策略,重写距离量子势能公式,定义相似性度量测度和相异性度量测度的新概念,提出了针对分类属性数据的量子聚类CQC算法,并对算法的聚类有效性进行了研究,通过同其它几个已有的算法的仿真实验比较,证明该算法是有效的、有一定的可扩展性,算法的一些性能优于已有的其它几个算法。
The distributive distribution among the samples of the classification attribute data is unbalanced, the distribution of the samples is independent of the spatial distance and the distribution of the particles in the quantum mechanics is determined by the energy, and the characteristics of the particle distribution are unbalanced. Based on this, a new clustering strategy based on quantum clustering QC algorithm is proposed to determine the clustering strategy of the cluster centers, to rewrite the quantum potential energy formula, to define a new concept of measure of similarity measure and measure of dissimilarity measure. Algorithm and the clustering effectiveness of the algorithm are studied. Compared with the simulation experiments of several other existing algorithms, the algorithm is proved to be effective and scalable. Some of the performance of the algorithm is better than the existing ones Several other algorithms.