University of Hertfordshire

Removing redundant features via clustering: preliminary results in mental task separation

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

Documents

  • IkFS

    Accepted author manuscript, 216 KB, PDF document

  • Renato Cordeiro De Amorim
  • Boris Mirkin
View graph of relations
Original languageEnglish
Title of host publicationLooking into the Future of Creativity and Decision Support Systems
Subtitle of host publicationProceedings of the 8th International Conference on Knowledge, Information and Creativity Support Systems
EditorsAndrzej M.J. Skulimowski
PublisherProgress & Business Publishers
Pages63-72
ISBN (Electronic)978-83-912831-8-9
ISBN (Print)978-83-912831-6-5
Publication statusPublished - 2013
EventThe 8th International Conference on Knowledge, Information and Creativity Support Systems - Krakow, Poland
Duration: 7 Nov 20139 Nov 2013

Conference

ConferenceThe 8th International Conference on Knowledge, Information and Creativity Support Systems
CountryPoland
CityKrakow
Period7/11/139/11/13

Abstract

Recent clustering algorithms have been designed to take into account the degree of relevance of each feature, by automatically calculating their weights. However, as the tendency is to evaluate each feature at a time, these algorithms may have difficulties dealing with features containing similar information. Should this information be relevant, these algorithms would set high weights to all such features instead of removing some due to their redundant nature.
In this paper we introduce an unsupervised feature selection method that targets redundant features. Our method clusters similar features together and selects a subset of representative features for each cluster. This selection is based on the maximum information compression index between each feature and its respective cluster centroid. We empirically validate out method by comparing with it with a popular unsupervised feature selection on three EEG data sets. We find that ours selects features that produce better cluster recovery, without the need for an extra user-defined parameter

ID: 9822739