改进的蚂蚁聚类算法*

提出了一种改进的基于对称点距离的蚂蚁聚类算法。该算法不再采用Euclidean距离来计算类内对象的相似性，而是使用新的对称点距离来计算相似性，在处理带有对称性质的数据集时，可以有效地识别给定数据集的聚类数目和合适的划分。在该算法中，用人工蚂蚁代表数据对象，根据算法给定的聚类规则来寻找最合适的聚类划分。最后用本算法与标准的蚂蚁聚类算法分别对不同的数据集进行了聚类实验。实验结果证实了算法的有效性。

Improved ant clustering algorithm

This paper proposed an improved ant clustering algorithm based on point symmetry distance. Assignment of points to different clusters was done based on point symmetry distance rather than the traditional Euclidean distance. It could defect the number of clusters and the proper partitions from data set when data sets possess the property of symmetry. In the algorithm, each ant represented a data object. It would decide its next moving position according to clustering rules. Comparing the standard ant clustering algorithm, it demonstrated the effectiveness of point symmetry-based ant clustering algorithm for different data sets.