Clustering is a somewhat confusing topic theoretically. In large part this is because there are many different kinds of clustering problems. In addition, the true goals in clustering are often difficult to measure, making the task seem not well-defined and under-specified. In this note we discuss one approach from [6] to theoretically formulating a certain broad class of clustering problems, an...

We propose a clustering method that considers non-rigid alignment of samples. The motivation for such a clustering is training of object detectors that consist of multiple mixture components. In particular, we consider the deformable part model (DPM) of Felzenszwalb et al., where each mixture component includes a learned deformation model. We show that alignment based clustering distributes the...

Medoids-based fuzzy relational clustering generates clusters of objects based on relational data, which records pairwise similarity or dissimilarities among objects. Compared with single-medoid based approaches, multiple-weighted medoids has shown superior performance in clustering. In this paper, we present a new version of fuzzy relational clustering in this family called fuzzy clustering wit...

Co-clustering or simultaneous clustering of rows and columns of two-dimensional data matrices, is a data mining technique with various applications such as text clustering and microarray analysis. Most proposed co-clustering algorithms work on the data matrices with special assumptions and they also assume the existence of a number of mutually exclusive row and column clusters, but it is believ...

Spectral clustering is a family of methods to find K clusters using the eigenvectors of a matrix. Typically, this matrix is derived from a set of pairwise similarities Sij between the points to be clustered. This task is called similarity based clustering, graph clustering, or clustering of diadic data. One remarkable advantage of spectral clustering is its ability to cluster “points” which are...

In this paper, we introduce a threshold free approach, motivated from Chinese Restaurant Process, for the purpose of cognate clustering. We show that our approach yields similar results to a linguistically motivated cognate clustering system known as LexStat. Our Chinese Restaurant Process system is fast and does not require any threshold and can be applied to any language family of the world.

Repetitive DNA elements were proposed to coordinate chromatin folding and interaction in chromosomes by their intrinsic homology-based clustering ability. A recent analysis of the data sets from chromosome-conformation-capture experiments confirms the spatial clustering of DNA repeats of the same family in the nuclear space, and thus provides strong new support for the CORE theory.