Directly Solving the Original Ratiocut Problem for Effective Data Clustering

Jing Li, Feiping Nie, Xuelong Li

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

9 Scopus citations

Abstract

This paper focuses on the original RatioCut problem, which is one of the most representative clustering paradigms. The RatioCut criterion looks for a partition of the graph to achieve the mincut cost while keeping each partition reasonably large. This well-known problem is NP hard and its relaxed form has been widely used in the past several decades. However, the relaxed RatioCut usually suffers two problems: not satisfactory stable clustering performance, and undesired two-stage optimization. In this work, we solve the original RatioCut problem by learning a new similarity matrix which has as many connected components as the cluster number, so that the original RatioCut constraint can be directly satisfied. An easily implemented algorithm is derived to iteratively optimize the proposed method. Experimental results on various real-world benchmark datasets exhibit the effectiveness of the proposed method to solve the RatioCut problem.

Original languageEnglish
Title of host publication2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages2306-2310
Number of pages5
ISBN (Print)9781538646588
DOIs
StatePublished - 10 Sep 2018
Event2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Calgary, Canada
Duration: 15 Apr 201820 Apr 2018

Publication series

NameICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Volume2018-April
ISSN (Print)1520-6149

Conference

Conference2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018
Country/TerritoryCanada
CityCalgary
Period15/04/1820/04/18

Keywords

  • Discrete constraint conditions
  • Iterative optimization
  • RatioCut clustering

Fingerprint

Dive into the research topics of 'Directly Solving the Original Ratiocut Problem for Effective Data Clustering'. Together they form a unique fingerprint.

Cite this