Analytical and Experimental Study of Filter Feature Selection Algorithms for High-dimensional Datasets
- DOI
- 10.2991/.2013.42How to use a DOI?
- Keywords
- feature selection, filter strategy, high-dimensional datasets, supervised classification
- Abstract
In this paper a new taxonomy for feature selection algorithms created for high-dimensional datasets is proposed. Also, several selectors are described, analyzed and evaluated. It was observed that the Cfs-SFS algorithm reached the best solutions in most of the cases. Nevertheless, its application in very high-dimensional datasets is not recommended due to its computational cost. Cfs-BARS, Cfs-IRU and MRMR algorithms have similar results to those of Cfs-SFS, but in a relatively lesser time. The INTERACT algorithm gets good solutions too, but its computational cost is higher if compared to the above mentioned. On the other hand, the QPFS and FSBMC algorithms reached the worst solutions.
- Copyright
- © 2013, the Authors. Published by Atlantis Press.
- Open Access
- This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).
Cite this article
TY - CONF AU - Adrian Pino AU - Carlos Morell PY - 2013/10 DA - 2013/10 TI - Analytical and Experimental Study of Filter Feature Selection Algorithms for High-dimensional Datasets BT - Proceedings of the Fourth International Workshop on Knowledge Discovery, Knowledge Management and Decision Support PB - Atlantis Press SP - 339 EP - 349 SN - 1951-6851 UR - https://doi.org/10.2991/.2013.42 DO - 10.2991/.2013.42 ID - Pino2013/10 ER -