Search academic texts

Information × Registration Number 0219U100513, 0118U003169 , R & D reports Title Methods of intellectual processing and analysis of large data based on deep neural networks popup.stage_title Head Sаchenko Anаtolii O., Доктор технічних наук Registration Date 15-02-2019 Organization Ternopil National Economic University popup.description2  The object of the study - the processes of intelligent processing and analysis of large data using deep neural networks. The goal is to develop methods and tools aimed at increasing the efficiency and productivity of intelligent processing and analysis of large data based on deep neural networks. Research methods - methods of system analysis, mathematical statistics, intelligent data analysis, large data, theory of pattern recognition, the theory of artificial neural networks, neural computing. The results of the study and their novelty: improved: the method of compressing large data on an example of network traffic parameters, which, unlike the known, is based on the use of a deep autoreconder neural network to reduce the size of the analyzed information, allows for the nonlinear nature of the neural elements to be taken into account and is characterized by minimization of the mean square error of information restoration, which made it possible to avoid the main shortcomings of existing methods and increase the reliability of the results; the method of classificationand large data on the example of attacks on information telecommunication networks using deep neural networks, which, unlike the known, allows to analyze unstructured data without their preprocessing, which allowed to improve the speed and made it possible to use in systems for detecting intrusions in real time; neural network image recognition method based on pre-processing of the image, which simplifies the localization of its individual parts and further recognition of localized blocks with the help of a deep convolutional neural network and allows to recognize low-resolution images. Product Description popup.authors popup.nrat_date 2020-04-02 Close