CLASSIFICATION, IDENTIFICATION AND TESTING OF PEAR GENOTYPES APPLYING ASK-ANALYSIS METHOD

The paper studied ask-analysis application and its software tools – intellectual systems "Eidos" for classifying, identifying and testing the cultivars and hybrids available in the collection and for establishing the relationships between pear morphological characteristics (size and shape of leaves, size, weight, shape of fruit and its index). The degree of similarity and differences between the indicators characterizing genotypes was studied. 3 statistical and 7 system-cognitive models were created, in which generalized images of classes were formed for quantitative and qualitative assessment of the complex of morphological features in pear genotypes. The most reliable model was INF4 with the integral criterion "Knowledge resonance", which was used to solve the forecasting problems (diagnostics, classification, recognition, identification).