STRONG ARTIFICIAL INTELLIGENCE: IMPROVING THE QUALITY OF NEURAL NETWORK SOLUTIONS WITH THE TRANSITION TO PROCESSING INPUT DATA OF A LARGE VOLUME 

Alexander I. Ivanov, Doctor of technical sciences, associate professor, consultant, Penza Research Electrotechnical Institute (9 Sovetskaya street, Penza, Russia), E-mail: ivan@pniei.penza.ru
Igor A. Kubasov, Doctor of technical sciences, associate professor, chief researcher of the FKU NPO "STIS" of the Ministry of Internal Affairs of Russian Federation, professor of sub-department of information technologies, Academy of management of the Ministry of internal affairs of the Russian Federation (8 Zoi i Aleksandra Kosmodem'yanskikh street, Moscow, Russia), E-mail: igorak@list.ru 

004.838.2 

10.21685/2307-4205-2021-1-1 

Background. The article examines the ways to solve the actual problem of translating applications of "weak" artificial intelligence into applications of "strong" artificial intelligence, used in the interests of improving the reliability and quality of complex technical systems. The possibility of improving the quality of neural network solutions with the transition to processing large-volume input data is justified.
Materials and methods. The influence of the number of input and output dimensions of an artificial neural network on the quality of decisions made is estimated. The solution of the problem of neural network analysis by means of symmetrization by correlation coefficients of input data is proposed.
Results. When planning a combination of the use of natural and artificial intelligence to improve the reliability and quality of complex technical systems, a quantitative assessment of the correlation between the categories of "big data" and "strong" artificial intelligence is proposed. The article presents a simple hypothesis of linking these two significant categories.
Conclusion. An estimate of the probability level of errors characteristic of a "strong" human expert when accessing real data of bounded dimension n is presented. Recommendations are given for achieving the desired level of reducing the probability of errors or setting the desired level of increasing the volume of input data taken into account. As a result, a forecast of how these two parameters that are subject to regulation (planning) are related to each other is obtained. 

artificial intelligence, neural networks, quality of neural network solutions, big data