Article 6222

Title of the article

DECISION TREES BASED ON MEMRISTOR TECHNOLOGY 

Authors

Anton Yu. Dorosinskiy, Candidate of technical sciences, general manager, Scientific and production enterprise «Sonar» (1V Tsentralnaya street, Penza, Russia), E-mail: antik_r13@mail.ru
Oleg V. Prokofev, Candidate of technical sciences, associate professor, associate professor of the sub-department of informational technologies and systems, Penza State Technological University (1a/11 Baidukov's passage /Gagarina street, Penza, Russia), E-mail: prokof_ow@mail.ru
Marina A. Linkova, Master degree student, Penza State Technological University (1a/11 Baidukov's passage /Gagarina street, Penza, Russia), E-mail: m__linkova@mail.ru
Irina Yu. Semochkina, Candidate of technical sciences, associate professor, associate professor of the sub-department of informational technologies and systems, Penza State Technological University (1a/11 Baidukov's passage /Gagarina street, Penza, Russia), E-mail: ius1961@gmail.com 

Abstract

Background. Despite significant progress in neuroscience recently, understanding of the principles and mechanisms underlying complex brain functions and cognition remains incomplete. Modeling and physical implementation of the decision-making center, similar to the natural processes of the brain, is a means of building a cyber-physical system for a wide range of applied tasks related to decision support. Matherials and methods. The discovery of memristors, the identification of the technical possibility of memorizing and storing analog information serve as a technological platform for creating decision trees. Neurons involved in information processing and data storage, dynamically establishing connections among themselves, are also implemented with voltage-controlled memristors. Neurobiology methods, network models and methods for calculating microelectronics circuits based on memristors are used here in a single complex for memorizing, processing information and making decisions. Results. The advantages that are being discovered allow not only to use information storage devices of higher capacity, replacing traditional flash memory, but also to use decision trees with branching nodes implemented at the molecular level. In conclusion, the inference is substantiated about the possibility of creating a new generation of decision support systems combining modern visualization tools of trees and the use of decision libraries. Conclusions. Thus, it is possible to integrate the advantages of an advisory, explanatory system and an expert system that independently develops a solution if a task in the subject area requires it. The development of microelectronic memristors makes it possible to develop intelligent decision support systems of a new generation that simulate the biological processes of the human brain, in which the processes of learning, creating, memorizing and storing decision trees are implemented using a single technological and circuit-based base. 

Key words

decision trees, memristors, decision support 

Download PDF
For citation

Dorosinskiy A.Yu., Prokofev O.V., Linkova M.A., Sеmochkina I.Yu. Decision trees based on memristor technology. Nadezhnost' i kachestvo slozhnykh sistem = Reliability and quality of complex systems. 2022;(2):53–60. (In Russ.). doi:10.21685/2307-4205-2022-2-6 

 

Дата создания: 01.07.2022 09:01
Дата обновления: 01.07.2022 09:29