Exact Learning and Data Compression with a Local Asymmetrically Weighted Metric

This paper introduces a new local asymmetric weighting scheme for the nearest neighbour classification algorithm. It is shown both with theoretical arguments and computer experiments that good compression rates can be achieved outperforming the accuracy of the standard nearest neighbour classification algorithm and obtaining almost the same accuracy as the k-NN algorithm with k optimised in each data set. Moreover, the learning procedure, based on reinforcement, is quite robust against suboptimal choices of the reinforcement, punishment and compression paramenter