Full Citation
Title: A Feature Subset Selection Algorithm Automatic Recommendation Method
Citation Type: Journal Article
Publication Year: 2013
ISBN:
ISSN:
DOI:
NSFID:
PMCID:
PMID:
Abstract: Many feature subset selection (FSS) algorithms have been proposed, but not all of themare appropriate for a given feature selection problem. At the same time, so far there israrely a good way to choose appropriate FSS algorithms for the problem at hand. Thus,FSS algorithm automatic recommendation is very important and practically useful. Inthis paper, a meta learning based FSS algorithm automatic recommendation method ispresented. The proposed method first identifies the data sets that are most similar to theone at hand by thek-nearest neighbor classication algorithm, and the distances amongthese data sets are calculated based on the commonly-used data set characteristics. Then,it ranks all the candidate FSS algorithms according to their performance on these similardata sets, and chooses the algorithms with best performance as the appropriate ones.The performance of the candidate FSS algorithms is evaluated by a multi-criteria metricthat takes into account not only the classification accuracy over the selected features, butalso the runtime of feature selection and the number of selected features. The proposedrecommendation method is extensively tested on 115 real world data sets with 22 well-known and frequently-used different FSS algorithms for five representative classifiers. Theresults show the effectiveness of our proposed FSS algorithm recommendation method
User Submitted?: No
Authors: Zhou, Yuming; Zhang, Xueying; Song, Qinbao; Sun, Heli; Xu, Baowen; Wang, Guangtao
Periodical (Full): Journal of Artificial Intelligence Resea
Issue:
Volume: 47
Pages: 1-34
Data Collections: IPUMS USA
Topics: Methodology and Data Collection
Countries: