Abstract. We introduce a hybrid approach to magnetic resonance image segmentation using unsupervised clustering and the rules derived from approximate decision reducts. We utilize ...
Abstract. A new methodology that structures the semantics of a collection of documents into the geometry of a simplicial complex is developed. A simplicial complex is topologically...
The segmentation of time-series is a constrained clustering problem: the data points should be grouped by their similarity, but with the constraint that all points in a cluster mus...
We present a novel multiscale clustering algorithm inspired by algebraic multigrid techniques. Our method begins with assembling data points according to local similarities. It us...
We consider clustering as computation of a structure of proximity relationships within a data set in a feature space or its subspaces. We propose a data structure to represent suc...