论文信息 - Scalability analysis of lter-based methods for feature selection

Scalability analysis of lter-based methods for feature selection

Researchers in machine learning are now interested not only in accuracy but also in scalability of methods. Although scalability of learning algorithms is a trending issue, scalability of feature selection methods has not received the same amount of attention. In this research, a preliminary attempt to study the scalability of three well-known filterbased feature selection methods will be done. For this sake, several new measures are introduced, based not only in accuracy but also in execution time and stability and the results will be presented according to them.

[1] Ludmila I. Kuncheva,et al. A stability index for feature selection , 2007, Artificial Intelligence and Applications.

[2] L. A. Belanche,et al. Review and Evaluation of Feature Selection Algorithms in Synthetic Problems , 2011, 1101.2320.

[3] Huan Liu,et al. Efficient Feature Selection via Analysis of Relevance and Redundancy , 2004, J. Mach. Learn. Res..

[4] Huan Liu,et al. Searching for Interacting Features , 2007, IJCAI.

[5] Sören Sonnenburg,et al. Optimized Cutting Plane Algorithm for Large-Scale Risk Minimization , 2009, J. Mach. Learn. Res..

[6] Masoud Nikravesh,et al. Feature Extraction - Foundations and Applications , 2006, Feature Extraction.

[7] Huan Liu,et al. Consistency-based search in feature selection , 2003, Artif. Intell..

[8] Leo Breiman,et al. Classification and Regression Trees , 1984 .

[9] Mark A. Hall,et al. Correlation-based Feature Selection for Machine Learning , 2003 .

[10] Ron Kohavi,et al. Irrelevant Features and the Subset Selection Problem , 1994, ICML.

[11] Huan Liu,et al. Spectral Feature Selection for Data Mining , 2011 .