Grynkevych G. Information technology for building networks of the Internet of things based on a distributed architecture

As a result of dissertation researches the important scientific and applied problem concerning development of information technology for building networks of the Internet of things based on a distributed architecture. For the first time the methodological bases of construction of the Internet of Things on the basis of interconnected set of model and methods based on the theory of machine learning taking into account the distribution of server resources in the learning process online optimization of finite IoT device computation time, background aggregate error minimization in distributed learning with reinforcement, and automated development and deployment of a distributed machine learning architecture for IoT devices. Developed information technology for building IoT networks based on distributed architecture, the scientific novelty of which is that it is based on the model of automated development and deployment of machine learning architecture and methods: machine learning software agents IoT-network with adaptive communication mechanism ; distribution of server resources in the process of training software agents of the IoT network; modelless online optimization of computational time of end IoT devices; background aggregate minimization of errors in distributed learning with the reinforcement of disparate software agents IoT-network. The proposed information technology allows to increase the efficiency of operation and management of services of modern IoT networks, built on a distributed architecture.