Number of the records: 1
Improving feature selection process resistance to failures caused by curse-of-dimensionality effects
- 1.
SYSNO ASEP 0368741 Document Type J - Journal Article R&D Document Type Journal Article Subsidiary J Článek ve WOS Title Improving feature selection process resistance to failures caused by curse-of-dimensionality effects Author(s) Somol, Petr (UTIA-B) RID
Grim, Jiří (UTIA-B) RID, ORCID
Novovičová, Jana (UTIA-B)
Pudil, P. (CZ)Number of authors 4 Source Title Kybernetika. - : Ústav teorie informace a automatizace AV ČR, v. v. i. - ISSN 0023-5954
Roč. 47, č. 3 (2011), s. 401-425Number of pages 25 s. Language eng - English Country CZ - Czech Republic Keywords feature selection ; curse of dimensionality ; over-fitting ; stability ; machine learning ; dimensionality reduction Subject RIV IN - Informatics, Computer Science R&D Projects 1M0572 GA MŠMT - Ministry of Education, Youth and Sports (MEYS) GA102/08/0593 GA ČR - Czech Science Foundation (CSF) CEZ AV0Z10750506 - UTIA-B (2005-2011) UT WOS 000293207900007 EID SCOPUS 83455221244 Annotation The purpose of feature selection in machine learning is at least two-fold – saving measurement acquisition costs and reducing the negative effects of the curse of dimensionality with the aim to improve the accuracy of the models and the classification rate of classifiers with respect to previously unknown data. Yet it has been shown recently that the process of feature selection itself can be negatively affected by the very same curse of dimensionality – feature selection methods may easily over-fit or perform unstably. Such an outcome is unlikely to generalize well and the resulting recognition system may fail to deliver the expectable performance. In many tasks, it is therefore crucial to employ additional mechanisms of making the feature selection process more stable and resistant the curse of dimensionality effects. In this paper we discuss three different approaches to reducing this problem. Workplace Institute of Information Theory and Automation Contact Markéta Votavová, votavova@utia.cas.cz, Tel.: 266 052 201. Year of Publishing 2012
Number of the records: 1