FIELD: automating activities in lending agencies including prediction of financial risks. SUBSTANCE: method is used to take decision on crediting potential borrowers upon their requests. Use is made of decision-taking algorithm reproducing functions of artificial neural network with neuron number equal to number of questions in crediting request plus two in intermediate layer and one neuron in output layer. System implementing proposed method has data input device, data output device, data processing device, credit rating computing device, weighting coefficient computing device, code assignment device, data storage device, comparison device, and credit limit computing device. EFFECT: enhanced capacity, efficiency, and reliability of business. 13 cl, 3 dwg, 1 tbl
