標題: The ANNIGMA-wrapper approach to fast feature selection for neural nets
作者: Hsu, CN
Huang, HJ
Schuschel, D
資訊工程學系
Department of Computer Science
關鍵字: curse of dimensionality;feature selection;neural networks (NNs);wrapper model
公開日期: 1-Apr-2002
摘要: This paper presents a novel feature selection approach for backprop neural networks (NNs). Previously, a feature selection technique known as the wrapper model was shown effective for decision trees induction. However, it is prohibitively expensive when applied to real-world neural net training characterized by large volumes of data and many feature choices. Our approach incorporates a weight analysis-based heuristic called artificial neural net input gain measurement approximation (ANNIGMA) to direct the search in the wrapper model and allows effective feature selection feasible for neural net applications. Experimental results on standard datasets show that this approach can efficiently reduce the number of features while maintaining or even improving the accuracy. We also report two successful applications of our approach in the helicopter maintenance applications.
URI: http://dx.doi.org/10.1109/3477.990877
http://hdl.handle.net/11536/28899
ISSN: 1083-4419
DOI: 10.1109/3477.990877
期刊: IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS
Volume: 32
Issue: 2
起始頁: 207
結束頁: 212
Appears in Collections:Articles


Files in This Item:

  1. 000174455700008.pdf