Исследование эффективности признаков для диагностики сосудистой патологии

A method of filtering the source data using clustering algorithm, to improve the quality of discriminant analysis of sample data, was proposed. To improve the quality of diagnostics we propose an algorithm for the informative features formation, using methods of discriminant analysis. As integral indicators of fundus vessels, a global set of geometric features was used. It is a complete characterization for vascular disease diagnosis.