S.D. Kulik - Dr. Sc. (Eng.), Professor, National Research Nuclear University «MEPHI»; Moscow State University of Psychology and Education (MSUPE). E-mail: sedmik@mail.ru

This paper briefly discusses the neural network and automated factographic information systems (AFIS). AFIS is the special class factographic systems. Method is proposed for develop of technical decision-making. This method contains the necessary neural networks. It is offered to use a special factographic retrieval and neural network. In paper questions of estimation of efficiency and estimated characteristics of efficiency of factgraphic retrieval by means of use neuronets algorithms are considered. Traditional methods of an esti-mation of efficiency do not allow to evaluate the effective-ness of factographic retrieval in AFIS. Briefly introduces the concept of building the AFIS. The variant of a possible scheme of AFIS for implementing it in practice are proposed. It is offered to use a special factographic information retrieval in criminalistics. This paper is about automated factographic information systems. Such AFIS can be used in different applied tasks for special fac-tographic retrieval and using it for retrieval special object of the retrieval system. That technology lets to increase efficiency of the factographic systems using search engines of the special class. Also it can be useful for systems that work with large vo-lumes of factographic data. AFIS consist several subsystems (analyze and decision-making unit, data preparing unit, factografic data retrieval, factografic database and other unit). Also AFIS have human friendly interface that helps users (operators) to work with it without special training. In this paper submitted descriptions of these subsystems and their blocs, methods and algorithms that using in different units and using of the neural networks. Author propose to use AFIS in criminalistics systems or in a special class of retrieval systems that use features in research process. Such retrieval system use the special algorithms and can be used for searching data of different types. Nowauthorconti-nueresearchesinthisscientificfield. The results of the researches were successfully protected by the various security documents (patents).

 

1. Galushkin A.I. Teorija nejjronnykh setejj: Uchebnoe posobie dlja vuzov. Nejjrokompjutery i ikh primenenie. Kn. 1. M.: IPRZHR. 2000. 416 s.
2. GalushkinA. I.NeuralNetworksTheory, SpringerBerlinHeidelbergNewYork. 2007. 396 p.
3. HaykinS.NeuralNetworks - AComprehensive Foundation, Second Edition, Pearson Education, Inc.1999. Reprint. 2005.
4. Nilsson Nils J.The Mathematical Foundations of Learning Machines. 2nd Edition, San Mateo, CA: Morgan Kaufmann Publishers Inc., 1990. 138 p. (Nilsson N. Learning machines. Mir, Moscow, 1967).
5. Specht D.F. Probabilistic neural networks // IEEE Transactions on Neural net-works. 1990. V.3. № 1. P. 109-118.
6. Kruglov I.A., Mishulina O.A,. Bakirov B.Quantile based decision making rule of the neural networks committee for ill-posed approximation problems // Neurocomputing. 2012. V. 96. P. 74-82.
7. Kulik S.D. Primenenie nejjronnykh setejj v avtomatizirovannykh faktograficheskikh informacionno-poiskovykh sistemakh // Nejjrokompjutery: razrabotka i primenenie. 2002. № 5-6. S. 3-12.
8. Fukunaga K. Introduction To Statistical Pattern Recognition - 2Nd - Elsevier Academic Press, San Diego, San Francisco, New York, Boston, London, Sydney, Tokyo. 1990. 592 p.
9. Salton G. Automatic information organization and Retrieval. New York: McGraw-Hill, 1968. 514 p.
10. Taha H.A. Operations research: an introduction (8th Edition).Upper Saddle River, New Jersey 07458. Pearson Prentice Hall. 2007. 813 p.
11. KolmogorovA.N. Osnovnyeponjatijateoriiverojatnostejj. M.: FAZIS. 1998. 142 s.
12. Feller W. An introduction to probability theory and its applications, V. 1, 3nd ed., New York: John Wiley & Sons. 1968. 509 p.
13. Ventcel E.S. Teorija verojatnostejj. M.: Vysshaja shkola. 2001. 575 s.
14. Kulik S.D. Ocenka ehffektivnosti poiskovykh operacijj// Prikladnaja informatika. 2014. № 6(54). S. 60-69.
15. Patent № 2208837 (RF). Ustrojjstvo dlja imitacionnogo modelirovanija znachenijj funkcii vykhoda avtomatizirovannojj faktograficheskojj informacionno-poiskovojj sistemy kriminalisticheskogo naznachenija (Rossija) / S.D. Kulik. Opubl. 20.07.2003.