Information Theoretic Clustering Using Minimum Spanning Trees
Information Theoretic Clustering Using Minimum Spanning Trees
复制标题
DOI：
10.1007/978-3-642-32717-9_21
复制
发表时间：
2012-08
期刊：
影响因子：
--
通讯作者：
Andreas C. Müller;Sebastian Nowozin;Christoph H. Lampert
中科院分区：
文献类型：
--
作者：
Andreas C. Müller;Sebastian Nowozin;Christoph H. Lampert
In this work we propose a new information-theoretic clustering algorithm that infers cluster memberships by direct optimization of a non-parametric mutual information estimate between data distribution and cluster assignment. Although the optimization objective has a solid theoretical foundation it is hard to optimize. We propose an approximate optimization formulation that leads to an efficient algorithm with low runtime complexity. The algorithm has a single free parameter, the number of clusters to find. We demonstrate superior performance on several synthetic and real datasets.
