NobleBlocks

Research indexJournal article

Автоматизированный синтез когнитивной модели на основе анализа больших данных и глубокого обучения

Райков

1 Jan 2018Open access

NobleIDark:/48914/w1/20260525/31764E23DOI10.17586/2541-9781-2018-2-103-111

Abstract

The cognitive models are created by team’s participants or experts from different subjects’ fields. They cannot quickly describe the problem situation with a high level of quality and integrity. Cognitive modeling takes into account denotative (formalized, material) and significative (cognitive, intellectual) semantics simultaneously. Еhe latter cannot be formalized. Cognitive modeling is a unique way to comprehensively reflect the features of the problem that has arisen. To ensure the integrity of the cognitive model, taking into account the fullest amount of information, as well as speeding up modeling and increasing the adequacy of models, this paper address the issue of performing automatic verification of cognitive models on the base of analyzing Big Data. It is also considered the possibility to perform automated synthesis of cognitive models. In the work the methods of deep learning, category theory and quantum semantics ideas are applied, as well as author's convergent approach to management and group decision making support. Experimentally the high level of quality of deep learning and accuracy of automatic recognition of elements of cognitive models are shown (93%). It can serve as a weighty guarantee of fruitfulness of offered solutions.

Discussion

Topics

More

Suggest a correction or report