Open AccessApplication prospects for vector time maps of cognitive images links
Author(s) -
E. A. Pustovoy,
O. A. Pustovaya
Publication year - 2019
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/643/1/012108
Subject(s) - memorization , computer science , scalability , flexibility (engineering) , artificial intelligence , cognition , process (computing) , artificial neural network , function (biology) , pattern recognition (psychology) , machine learning , mathematics , psychology , database , statistics , mathematics education , neuroscience , evolutionary biology , biology , operating system
Neural networks have a great advantage in the analysis of implicit dependencies, big data analysis, image recognition and other recognition systems. However, they do not perform the function of memorizing information in an explicit form, they cannot logically think like a person. Neural networks are not managed repositories of information about cognitive images. They are quite difficult to edit or create new structures based on existing trained models. They do not have the necessary flexibility and scalability in the process of learning and in use. We propose to create vector-time communication maps of cognitive images links, containing functions for memorizing individual cognitive images and their links, with the possibility of forming explicit centers and with the possibility of scaling.