一种基于密度的增量 k-means 聚类算法研究

介绍了 k -means 和 DBSCAN 聚类算法的基本原理和优缺点，针对传统聚类算法无法有效处理高维混合属性数据集的问题，对原有的数据归一化方法进行改进，在 k -means 和 DBSCAN 聚类算法的基础之上，结合增量聚类的思想和数据之间相异度的计算方法，提出了基于密度的增量 k -means 聚类算法，有效处理具有高维混合属性的数据集，改进了数据相异度的计算方法。

The Research on Incremental k-means Clustering Algorithm Based on Density

The basic principle,advantages and disadvantages of the k-means and DBSCAN clustering algo-rithms are firstly introduced in this article.Due to the fact that the traditional clustering algorithm can not effectively deal with high dimensional mixed datasets,it improves the original data normalization method. Based on the k-means and DBSCAN clustering algorithms,and the combination of incremental clustering and data dissimilarity in data calculation,it also puts forward an incremental k-means clustering algorithm based on density,which has improved the data dissimilarity algorithm.