期刊名称：Visnyk NTUU KPI : Informatics, Operation and Computer Science
印刷版ISSN：0135-1729
出版年度：2008
卷号：48
出版社：National Technical University of Ukraine
摘要：The paper deals with offers how to improve the using of neuronet technologies for remote multi-user systems subscribers identification. Usage of the perceptron model as the ambiguous reversible functional transformation allows to accelerate the identification process based on “zero knowledge” concepts. New ways of speeding up the perceptron model are exposed. What the article brings out is the new calculation scheme. Its efficiency is discovered both analytically and experimentally in order to grade up the identification performance.
