Scientific journal

ISSN 1814-2400

INFORMATION SCIENCE AND CONTROL SYSTEMS

Lapko A. V., Lapko V. A.

SYNTHESIS AND ANALYSIS OF THE COLLECTIVE OF NONPARAMETRIC DECISION FUNCTIONS IN THE TWO-OPTIONAL PROBLEMS OF PATTERN RECOGNITION

The authors of the article proposed a technique for synthesizing a collective of nonparametric equations of separating surfaces in the two-optional problem of pattern recognition, based on the decomposition of the training sample by its volume. The asymptotic properties of the collective of decision functions are investigated here. There was also made a comparison of its properties with the traditional nonparametric estimation of the equation of the separating surface between the classes.

Keywords: pattern recognition, nonparametric statistics, large samples, collectives of decision functions, asymptotic properties