Elements of system analysis of programs using convolutional neural networks

Kulik Sergey Dmitrievich, Shtanko Alexander Nikolaevich

This article deals with the analysis of programs in the description of which convolutional neural networks are mentioned. Only programs protected by titles of protection of the Russian Federation were analyzed. The search was carried out among more than 140 thousand documents. A total of 227 programs and one database were found and reviewed. The areas of application of the found programs were identified. Particular attention in the search and analysis was paid to recognition for forensics (security elements of documents) and for medicine (psoriasis and PUVA therapy). The results obtained can be used by developers of neural network algorithms for automated factographic information systems.

Keywords:system analysis, computer program, convolutional neural networks, intellectual property item.