Likelihood based hierarchical clustering and network topology identification

This paper develops a new method for hierarchical clustering based on a generative dendritic cluster model. The objects are viewed as being generated through a tree structured refinement process. In certain problems, this generative model naturally captures the physical mechanisms responsible for relationships among objects, for example, in genetic studies and network topology identification. The networking problem is examined in some detail, to illustrate the new clustering method. In general, the generative model is not representative of actual physical mechanisms, but it nonetheless provides a means for dealing with errors in the similarity matrix, simultaneously promoting two desirable features in clustering: intra-class similarity and inter-class dissimilarity