A computational demand optimization aide for nearest-neighbor-based decision systems

An approach to the problem of computational demand minimization, via optimal subset selection from a given training data set, in the context of nearest-neighbor-based decision systems is presented. The approach attempts to obtain a consistent subset which in addition is minimal in its size. This minimal consistent subset selection leads to a unique solution irrespective of the initial order of presentation of the data. The consistency property is assured at every iteration. The samples are selected in the order of significance of their contribution for enabling the consistency property. This provides insight into the relative significance of the samples in the training set. Numerical examples are included to illustrate the methodology