稳定的聚类很好The stability of a good clustering |
|
课程网址: | http://videolectures.net/mlss06tw_meila_sgc/ |
主讲教师: | Marina Meila |
开课单位: | 华盛顿大学 |
开课时间: | 2007-02-25 |
课程语种: | 英语 |
中文简介: | 如果我们找到了数据集X的“好”聚类C,我们能否证明C与该数据集的(未知)最佳聚类C *相差不远? 也许令人惊讶的是,这个问题的答案有时是肯定的。 我们可以显示两个聚类代价函数的距离(C,C *)的界限:标准化剪切和K均值聚类的平方距离成本。 当数据X对给定成本允许“良好”聚类时,存在这些界限。 |
课程简介: | If we have found a "good" clustering C of data set X, can we prove that C is not far from the (unknown) best clustering C* of this data set? Perhaps surprisingly, the answer to this question is sometimes yes. We can show bounds on the distance( C, C* ) for two clustering cost functions: the Normalized Cut and the squared distance cost of K-means clustering. These bounds exist in the case when the data X admits a "good" clustering for the given cost. |
关 键 词: | 数据集; 最佳聚类; 聚类代价函数 |
课程来源: | 视频讲座网 |
最后编审: | 2019-07-16:cjy |
阅读次数: | 55 |