Membership embedding space approach and spectral clustering

Authors:
Stefano Rovetta;Francesco Masulli;Maurizio Filippone
Affiliations:
Department of Computer and Information Sciences, University of Genova and CNISM, Genova, Italy;Department of Computer and Information Sciences, University of Genova and CNISM, Genova, Italy;Department of Computer and Information Sciences, University of Genova and CNISM, Genova, Italy
Venue:
KES'07/WIRN'07 Proceedings of the 11th international conference, KES 2007 and XVII Italian workshop on neural networks conference on Knowledge-based intelligent information and engineering systems: Part III
Year:
2007

Citing 8
Cited 0

Normalized Cuts and Image Segmentation

IEEE Transactions on Pattern Analysis and Machine Intelligence
Pattern Recognition with Fuzzy Objective Function Algorithms

Pattern Recognition with Fuzzy Objective Function Algorithms
Redefining Clustering for High-Dimensional Applications

IEEE Transactions on Knowledge and Data Engineering
A New Cluster Isolation Criterion Based on Dissimilarity Increments

IEEE Transactions on Pattern Analysis and Machine Intelligence
A generalized kernel approach to dissimilarity-based classification

The Journal of Machine Learning Research
Kernel Methods for Pattern Analysis

Kernel Methods for Pattern Analysis
Shared farthest neighbor approach to clustering of high dimensionality, low cardinality data

Pattern Recognition
A tutorial on spectral clustering

Statistics and Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

The data representation strategy termed "Membership Embedding" is a type of similarity-based representation that uses a set of data items in an input space as reference points (probes), and represents all data in terms of their membership to the fuzzy concepts represented by the probes. The technique has been proposed as a concise representation for improving the data clustering task. In this contribution, it is shown that this representation strategy yields a spectral clustering formulation, and this may account for the improvement in clustering performance previously reported. Then the problem of selecting an appropriate set of probes is discussed in view of this result.