The sample properties evaluation for pattern recognition and intelligent diagnosis

Abstract

The problem of development of indicators characterizing quantitative the training sample properties for the problems of pattern recognition and intelligent diagnosis is solved. It includes such measures as a sample monotonicity, complexity, repetition, relative dimensionality, relative dependence approximation simplicity, relative inconsistency, evenness, class separability and compactness, integrated criteria of sample quality evaluation, sample and feature selection criteria. The using of offered criterions in practice allows to automatize the process of a construction, analysis and comparison of neural models for pattern recognition problem.