NONPARAMETRIC ALGORITHM OF AUTOMATIC CLASSIFICATION UNDER CONDITIONS OF LARGE-SCALE STATISTICAL DATA
A nonparametric algorithm for automatic classification of large-scale statistical data, based on probability density estimates of the Parsen type, is proposed. Its application makes it possible to isolate compact sets of points corresponding to unimodal fragments of probability density.
Keywords: nonparametric statistics, automatic classification, pattern recognition, probability density, large samples
