Deep neural networks application in next-generation intelligent computer systems

Abstract

In the article, an approach to building hybrid next-generation intelligent computer systems (NGICS) based on the integration of pre-trained models of deep neural networks and logical models developed using the OSTIS technology is proposed. To reduce the requirements for the size of the training dataset, the authors propose an alternative method for pre-training deep models. To achieve the interpretability of neural network models, the authors used methods from the Explainable AI (XAI) field.