This paper proposes a kNN model-based feature selection method aimed at improving the efficiency and effectiveness of the ReliefF method by: (1) using a kNN model as the starter selection, aimed at choosing a set of more meaningful representatives to replace the original data for feature selection; (2) integration of the Heterogeneous Value Difference Metric to handle heterogeneous applications – those with both ordinal and nominal features; and (3) presenting a simple method of difference function calculation based on inductive information in each representative obtained by kNN model. We have evaluated the performance of the proposed kNN model-based feature selection method on toxicity dataset Phenols with two different endpoints. Experimental results indicate that the proposed feature selection method has a significant improvement in the classification accuracy for the trial dataset.
ER -
%0 Journal Article
%1 keyhere
%A Guo, Gongde
%A Neagu, Daniel
%A Cronin, Mark T.D.
%D 2005
%J Pattern Recognition and Data Mining
%K unread
%P 410--419
%T Using kNN Model for Automatic Feature Selection
%U http://dx.doi.org/10.1007/11551188_44
%X This paper proposes a kNN model-based feature selection method aimed at improving the efficiency and effectiveness of the ReliefF method by: (1) using a kNN model as the starter selection, aimed at choosing a set of more meaningful representatives to replace the original data for feature selection; (2) integration of the Heterogeneous Value Difference Metric to handle heterogeneous applications – those with both ordinal and nominal features; and (3) presenting a simple method of difference function calculation based on inductive information in each representative obtained by kNN model. We have evaluated the performance of the proposed kNN model-based feature selection method on toxicity dataset Phenols with two different endpoints. Experimental results indicate that the proposed feature selection method has a significant improvement in the classification accuracy for the trial dataset.
ER -
@article{keyhere,
abstract = {This paper proposes a kNN model-based feature selection method aimed at improving the efficiency and effectiveness of the ReliefF method by: (1) using a kNN model as the starter selection, aimed at choosing a set of more meaningful representatives to replace the original data for feature selection; (2) integration of the Heterogeneous Value Difference Metric to handle heterogeneous applications – those with both ordinal and nominal features; and (3) presenting a simple method of difference function calculation based on inductive information in each representative obtained by kNN model. We have evaluated the performance of the proposed kNN model-based feature selection method on toxicity dataset Phenols with two different endpoints. Experimental results indicate that the proposed feature selection method has a significant improvement in the classification accuracy for the trial dataset.
ER -},
added-at = {2008-10-17T10:57:35.000+0200},
author = {Guo, Gongde and Neagu, Daniel and Cronin, Mark T.D.},
biburl = {https://www.bibsonomy.org/bibtex/29a43a1255896104679f67db7333593ac/rgolombe},
description = {SpringerLink - Book Chapter},
interhash = {47f9bf20813ed7a00e4ba91b9defa351},
intrahash = {9a43a1255896104679f67db7333593ac},
journal = {Pattern Recognition and Data Mining},
keywords = {unread},
pages = {410--419},
timestamp = {2008-10-17T10:57:35.000+0200},
title = {Using kNN Model for Automatic Feature Selection},
url = {http://dx.doi.org/10.1007/11551188_44},
year = 2005
}