Classifier Design Based on the Use of Nearest Neighbor Samples

A considerable amount of effort has been devoted to design a classifier in small training sample size situations. In this paper, we propose to design a nonparametric classifier based on the use of nearest neighbor samples. In the experiments, both the artificial and real data sets were used. The proposed classifier is compared with the 1-NN, k-NN, and Euclidean distance classifiers in terms of the error rate, in small training sample size situations. Experimental results show that the proposed classifier is very effective, even in practical situations.