高维数据的可视化和快速聚类算法

本文通过介绍一种用于高维数据的可视化方法，引入了可用于快速聚类的一种距离算法，该方法不仅具有鲁棒性而且有较低的计算复杂性O（n^1），最后我们将该方法用于金融数据立方体的聚类算法，主要用于挖掘庄家行为模式并作为是否存在操纵行为的依据。

Visualization of High-dimension Data and the Study of Quick-Clustering Algorithms

This paper gives an approach to a distance algorithms of which can be used into quick-clustering through a visual method of high-dimension data. The new method has robustness and lower computering complexity. At last, it is used into clustering algorithms of financial data cube, which is used to mine banker deed mode and whether is pursuant from manipulating deed or not.