USE OF MACHINE LEARNING TECHNIQUES FOR EXTRACTION OF ASSOCIATION RULES IN DATASETS OF PLANTS AND ANIMALS CONTAINING MOLECULAR GENETIC MARKERS ACCOMPANIED BY CLASSIFICATION OR PREDICTION USING FEATURES CREATED BY THESE ASSOCIATION RULES Russian patent published in 2017 - IPC G06N5/02 

FIELD: biotechnology.

SUBSTANCE: invention relates to method of prediction of presence of at least one target feature in plant. Plant genotype is determined by direct sequencing of DNA for at least one molecular genetic marker. Dataset containing set of variables is provided, wherein at least one of variables in dataset has value, presenting plant genotype (s) for molecular genetic marker(s). At least one rule of association of dataset is determined using one or more extraction algorithms of rules of association, wherein rule of association is a rule determining elements, which frequently appear together within dataset. Rule(s) of association is used to create one or more new variables for dataset. New variable(s) is added to dataset and used to predict presence of target features in plant.

EFFECT: technical result consists in increase of accuracy of predicting target features in plants.

41 cl, 4 tbl, 1 dwg