Search academic texts

Information × Registration Number 0207U008739, 0107U010204 , R & D reports Title Research of methods of entering aprioristic of the diagnostic information in structure artificial neural of networks popup.stage_title Head Lokazyuk V.M., Registration Date 20-12-2007 Organization Khmelnytskiy National University popup.description2 In accordance with the dynamics of development and expansion of purviews computer systems (СS), rise requirement to reliability of their functioning. It does the problem of development and introduction of more effective controls and diagnoses of СS actually, especially on the stage of their exploitation. Modern facilities of diagnoses of CS are using intellectual components, do not provide the sufficient level of universality even at the decision of the separate strictly specialized troubleshooting problems, therefore questions of their development and improvement are actually and require further researches. The purpose of research works is a decision of scientific issue of the day of increase of efficiency of diagnoses of CS on the stage of exploitation by research-and-development methods of bringing of a priori diagnostic information in the structure of artificial neuron networks, being the constituent of the systems of diagnoses of CS. It is necessary for complete realization of idea of project: toprobe the types of diagnostic information; to choose the types of diagnostic information, suitable for bringing in the structure of artificial neuron networks; to probe efficiency of involving of different types of diagnostic information for different architectures of artificial neuron networks. Introduction of results of scientific researches is foreseen at development of the systems of the technical diagnoses for enterprises, engaged in assembling and repair of the computer systems, and also enterprises which use a computer technique in the process of the production activity. Product Description popup.authors popup.nrat_date 2020-04-02 Close