Clustering classifiers learnt from local datasets based on cosine similarity

Kaikai Zhao, Einoshin Suzuki

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

In this paper we present a new method to measure the degree of dissimilarity of a pair of linear classifiers. This method is based on the cosine similarity between the normal vectors of the hyperplanes of the linear classifiers. A significant advantage of this method is that it has a good interpretation and requires very little information to exchange among datasets. Evaluations on a synthetic dataset, a dataset from the UCI Machine Learning Repository, and facial expression datasets show that our method outperforms previous methods in terms of the normalized mutual information.

Original languageEnglish
Title of host publicationFoundations of Intelligent Systems - 22nd International Symposium, ISMIS 2015, Proceedings
EditorsFloriana Esposito, Olivier Pivert, Stefano Ferilli, Zbigniew W. Raś, Mohand-Saïd Hacid
PublisherSpringer Verlag
Pages150-159
Number of pages10
ISBN (Print)9783319252513
DOIs
Publication statusPublished - 2015
Event22nd International Symposium on Methodologies for Intelligent Systems, ISMIS 2015 - Lyon, France
Duration: Oct 21 2015Oct 23 2015

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9384
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other22nd International Symposium on Methodologies for Intelligent Systems, ISMIS 2015
Country/TerritoryFrance
CityLyon
Period10/21/1510/23/15

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'Clustering classifiers learnt from local datasets based on cosine similarity'. Together they form a unique fingerprint.

Cite this