ПРОГНОЗУВАННЯ СТАНІВ І ЕФЕКТУ ПРИ ДОПОМІЖНИХ РЕПРОДУКТИВНИХ ТЕХНОЛОГІЯХ НА ОСНОВІ БАГАТОРІВНЕВИХ ОНТОЛОГІЙ

Abstract

<p>The problem of prediction on the basis ontological knowledge representation of assisted reproductive technologies is considered. A number of levels medical care for patients with infertility problems is identified. At each level the set of surviving in the memory elements of information given on it with a binary relation induced by associative links between the various elements, bijectively is assigned oriented connected graph G.</p><p>Forecasting is ensured by using a decision rule oriented on the computation of projected state indicator. It is calculated by comparing the transition probabilities of the initial states available to patient, in prospective with probability of transition from other states in the same analyzed element graph. Demonstrated possibility practical implementation of the suggested approach. The forecasting accuracy for the first period accumulation of data completely persuasive and is (88,2 ±4) % is shown.</p>