Clustering by latent dimensions

05/28/2018

∙

by   Shohei Hidaka, et al.

∙

0

∙

share

This paper introduces a new clustering technique, called dimensional clustering, which clusters each data point by its latent pointwise dimension, which is a measure of the dimensionality of the data set local to that point. Pointwise dimension is invariant under a broad class of transformations. As a result, dimensional clustering can be usefully applied to a wide range of datasets. Concretely, we present a statistical model which estimates the pointwise dimension of a dataset around the points in that dataset using the distance of each point from its n^th nearest neighbor. We demonstrate the applicability of our technique to the analysis of dynamical systems, images, and complex human movements.