Smoothness bias in relevance estimators for feature selection in regression

Alexandra Degeest, Michel Verleysen, Benoît Frénay

Research output: Contribution in Book/Catalog/Report/Conference proceedingConference contribution

Abstract

Selecting features from high-dimensional datasets is an important problem in machine learning. This paper shows that in the context of filter methods for feature selection, the estimator of the criterion used to select features plays an important role; in particular the estimators may suffer from a bias when comparing smooth and non-smooth features. This paper analyses the origin of such bias and investigates whether this bias influences the results of the feature selection process. Results show that non-smooth features tend to be penalised especially in small datasets.

Original languageEnglish
Title of host publicationArtificial Intelligence Applications and Innovations - 14th IFIP WG 12.5 International Conference, AIAI 2018, Proceedings
EditorsIlias Maglogiannis, Lazaros Iliadis, Vassilis Plagianakos
PublisherSpringer New York
Pages285-294
Number of pages10
ISBN (Print)9783319920061
DOIs
Publication statusPublished - 1 Jan 2018
Event14th IFIP WG 12.5 International Conference on Artificial Intelligence Applications and Innovations, AIAI 2018 - Rhodes, Greece
Duration: 25 May 201827 May 2018

Publication series

NameIFIP Advances in Information and Communication Technology
Volume519
ISSN (Print)1868-4238

Conference

Conference14th IFIP WG 12.5 International Conference on Artificial Intelligence Applications and Innovations, AIAI 2018
CountryGreece
CityRhodes
Period25/05/1827/05/18

Keywords

  • Feature selection
  • Filter methods
  • Mutual information
  • Noise variance
  • Smoothness

Fingerprint Dive into the research topics of 'Smoothness bias in relevance estimators for feature selection in regression'. Together they form a unique fingerprint.

  • Cite this

    Degeest, A., Verleysen, M., & Frénay, B. (2018). Smoothness bias in relevance estimators for feature selection in regression. In I. Maglogiannis, L. Iliadis, & V. Plagianakos (Eds.), Artificial Intelligence Applications and Innovations - 14th IFIP WG 12.5 International Conference, AIAI 2018, Proceedings (pp. 285-294). (IFIP Advances in Information and Communication Technology; Vol. 519). Springer New York. https://doi.org/10.1007/978-3-319-92007-8_25