An Efficient Pseudo Nearest Neighbor Classifier.

K-nearest neighbor (KNN) rule is a very simple and efficient non-parametric classification algorithm that is widely used in machine learning. In this paper, we proposed a attribute weighting local-mean pseudo nearest neighbor rule (AWLMPNN). The main difference of AWLMPNN and local mean-based pseudo nearest neighbor (LMPNN) is that they use attribute weighting distance and Euclidean distance to measure the distance between two samples, respectively. To illustrate the effectiveness of the proposed AWLMPNN method, extensive experiments on 30 real UCI data sets are conduced by comparing with four competing KNN-based methods. The experimental results show that the proposed AWLMPNN method is superior to other methods, especially in the case of high dimensional attributes with small sample size. [ABSTRACT FROM AUTHOR]