Comparison Between Filter Criteria for Feature Selection in Regression

Degeest, Alexandra;Verleysen, Michel;Frénay, Benoît
(2019) ICANN 2019 — Location: Munich (17.September.2019)

Files

Adegeest_ICANN2019.pdf
  • Open Access
  • Adobe PDF
  • 524.81 KB
Download

Details

Authors
  • Degeest, AlexandraUCLouvain
    Author
  • Verleysen, Michelorcid-logoUCLouvain
    Author
  • Frénay, BenoîtUNamur
    Author
Abstract
High-dimensional data are ubiquitous in regression. To obtain a better understanding of the data or to ease the learning process, reducing the data to a subset of the most relevant features is important. Among the different methods of feature selection, filter methods are popular because they are independent from the model. This paper focuses on which properties make a good filter criterion, in order to be able to select one among the numerous existing ones. Six properties are discussed and three relevance criteria are compared with respect to the aforementioned properties.
Affiliations

Citations

  • APA
  • Chicago
  • FWB

Degeest, A., Verleysen, M., & Frénay, B. (2019). Comparison Between Filter Criteria for Feature Selection in Regression. Lecture Notes in Computer Science. Published. ICANN 2019, Munich. https://hdl.handle.net/2078.5/125967 (Original work published 2019)