Dataset complexity can help to generate accurate ensembles of k-nearest neighbors Export

@inproceedings{citeulike:6055905, abstract = {Gene expression based cancer classification using classifier ensembles is the main focus of this work. A new ensemble method is proposed that combines predictions of a small number of k-nearest neighbor (k-NN) classifiers with majority vote. Diversity of predictions is guaranteed by assigning a separate feature subset, randomly sampled from the original set of features, to each classifier. Accuracy of k-NNs is ensured by the statistically confirmed dependence between dataset complexity, determining how difficult is a dataset for classification, and classification error. Experiments carried out on three gene expression datasets containing different types of cancer show that our ensemble method is superior to 1) a single best classifier in the ensemble, 2) the nearest shrunken centroids method originally proposed for gene expression data, and 3) the traditional ensemble construction scheme that does not take into account dataset complexity.}, author = {Okun, O. and Valentini, G.}, citeulike-article-id = {6055905}, citeulike-linkout-0 = {http://dx.doi.org/10.1109/IJCNN.2008.4633831}, citeulike-linkout-1 = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4633831}, day = {26}, doi = {10.1109/IJCNN.2008.4633831}, journal = {Neural Networks, 2008. IJCNN 2008. (IEEE World Congress on Computational Intelligence). IEEE International Joint Conference on}, keywords = {bioinformatics, dataset\_complexity, ensemble, nearest\_neighbor}, month = {September}, pages = {450--457}, posted-at = {2009-11-02 11:03:26}, priority = {2}, title = {Dataset complexity can help to generate accurate ensembles of k-nearest neighbors}, url = {http://dx.doi.org/10.1109/IJCNN.2008.4633831}, year = {2008} }

TY - CONF ID - citeulike:6055905 L3 - citeulike-article-id:6055905 N2 - Gene expression based cancer classification using classifier ensembles is the main focus of this work. A new ensemble method is proposed that combines predictions of a small number of k-nearest neighbor (k-NN) classifiers with majority vote. Diversity of predictions is guaranteed by assigning a separate feature subset, randomly sampled from the original set of features, to each classifier. Accuracy of k-NNs is ensured by the statistically confirmed dependence between dataset complexity, determining how difficult is a dataset for classification, and classification error. Experiments carried out on three gene expression datasets containing different types of cancer show that our ensemble method is superior to 1) a single best classifier in the ensemble, 2) the nearest shrunken centroids method originally proposed for gene expression data, and 3) the traditional ensemble construction scheme that does not take into account dataset complexity. JF - Neural Networks, 2008. IJCNN 2008. (IEEE World Congress on Computational Intelligence). IEEE International Joint Conference on EP - 457 TI - Dataset complexity can help to generate accurate ensembles of k-nearest neighbors SP - 450 KW - bioinformatics KW - dataset_complexity KW - ensemble KW - nearest_neighbor AU - Okun, O AU - Valentini, G PY - 2008/09/26/ UR - http://dx.doi.org/10.1109/IJCNN.2008.4633831 ER -