Bayesian interpretation to generalize adaptive mean shift algorithm

Research output: Contribution to journalArticlepeer-review

Abstract

The Adaptive Mean Shift (AMS) algorithm is a popular and simple non-parametric clustering approach based on Kernel Density Estimation. In this paper the AMS is reformulated in a Bayesian framework, which permits a natural generalization in several directions and is shown to improve performance. The Bayesian framework considers the AMS to be a method of obtaining a posterior mode. This allows the algorithm to be generalized with three components which are not considered in the conventional approach: node weights, a prior for a particular location, and a posterior distribution for the bandwidth. Practical methods of building the three different components are considered.

Original languageEnglish
Pages (from-to)3583-3592
Number of pages10
JournalJournal of Intelligent and Fuzzy Systems
Volume30
Issue number6
DOIs
Publication statusPublished - 2016 Apr 30

Keywords

  • Adaptive mean shift algorithm
  • kernel density estimation

ASJC Scopus subject areas

  • Statistics and Probability
  • Engineering(all)
  • Artificial Intelligence

Fingerprint

Dive into the research topics of 'Bayesian interpretation to generalize adaptive mean shift algorithm'. Together they form a unique fingerprint.

Cite this