Alpha shapes in kernel density estimation

arXiv2303.12213MaRDI QIDQ6509271FDOQ6509271

Erik Carlsson, John Gunnar Carlsson

Abstract: We define a filtered simplicial complex associated to the superlevel sets of a sum of weighted Gaussians kernels $f(x)$, with uniform scale parameter $h>0$. Regarding $f$ as a kernel density estimator of a data set $mathcalDsubsetmathbb{R}^m$, we obtain a method for filtering persistent homology by density in low or high dimensions. On the other hand, we also see that our construction can be highly visually descriptive.

Mathematics Subject Classification ID

Density estimation (62G07) Persistent homology and applications, topological data analysis (55N31) Topological data analysis (62R40)

This page was built for publication: Alpha shapes in kernel density estimation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6509271)