2006 | OriginalPaper | Buchkapitel
Feature Ranking Using an EDA-based Wrapper Approach
verfasst von : Yvan Saeys, Sven Degroeve, Yves Van de Peer
Erschienen in: Towards a New Evolutionary Computation
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Feature subset selection is an important pre-processing step for classification. A more general framework of feature selection is feature ranking. A feature ranking provides an ordered list of the features, sorted according to their relevance. Using such a ranking provides a better overview of the feature elimination process, and allows the human expert to gain more insight into the processes underlying the data. In this chapter, we describe a technique to derive a feature ranking directly from the estimated distribution of an EDA. As an example, we apply the method to the biological problem of acceptor splice site prediction, demonstrating the advantages for knowledge discovery in biological datasets with many features.