Weighted Agglomerative Clustering to Solve Normalized Cuts Problems

Giulio Genovese

2008

Abstract

A new agglomerative algorithm is introduced that can be used as a replacement for any partitioning algorithm that tries to optimize an objective function related to graph cuts. In particular, spectral clustering algorithms fall in this category. A new measure of similarity is introduced to show that the approach, although radically different from the one adopted in partitioning approaches, tries to optimize the same objective. Experiments are performed for the problem of image segmentation but the idea can be applied to a broader range of applications.

Download


Paper Citation


in Harvard Style

Genovese G. (2008). Weighted Agglomerative Clustering to Solve Normalized Cuts Problems . In Proceedings of the 8th International Workshop on Pattern Recognition in Information Systems - Volume 1: PRIS, (ICEIS 2008) ISBN 978-989-8111-42-5, pages 67-76. DOI: 10.5220/0001740100670076


in Bibtex Style

@conference{pris08,
author={Giulio Genovese},
title={Weighted Agglomerative Clustering to Solve Normalized Cuts Problems},
booktitle={Proceedings of the 8th International Workshop on Pattern Recognition in Information Systems - Volume 1: PRIS, (ICEIS 2008)},
year={2008},
pages={67-76},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001740100670076},
isbn={978-989-8111-42-5},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 8th International Workshop on Pattern Recognition in Information Systems - Volume 1: PRIS, (ICEIS 2008)
TI - Weighted Agglomerative Clustering to Solve Normalized Cuts Problems
SN - 978-989-8111-42-5
AU - Genovese G.
PY - 2008
SP - 67
EP - 76
DO - 10.5220/0001740100670076