Search academic texts

Information × Registration Number 0126U000515, R & D request Title Artificial intelligence methods and tools for analyzing, collecting, and processing information in decision-making systems Head Serhii O. Subbotin, Доктор технічних наук Registration Date 20-01-2026 Organization Zaporizhzhya Polytechnic National University popup.description1 The research proposes to solve the urgent problem of developing new methods and means of artificial intelligence for synthesizing models based on artificial intelligence and high-performance computing, which will accelerate the synthesis of diagnostic models and increase their level of logical transparency for use in civil and military decision-making systems. The project is dedicated to solving this problem, and its results will be applied in military and civil engineering, radio electronics and instrument making, medicine, etc. The project involves the development and research of methods for compact encoding of information about neural models to accelerate the synthesis process, the development and research of a method for synthesizing models based on a neuroevolutionary approach, the development and research of a parallel method for synthesizing models for organizing work with large amounts of data in diagnostic and forecasting systems deployed on the basis of high-performance computing systems, the development and research of a method for fine structural tuning of synthesized models, the development and research of a model of information technology for diagnosis and forecasting, as well as the creation of software based on the developed methods and models; experimental research of developed methods, models, and software tools by solving practical problems of diagnosis and forecasting. The goal of the project is to develop and study methods and tools for decision-making for data processing in systems for diagnosing and predicting the condition of objects based on computational intelligence and high-performance computing. popup.nrat_date 2026-01-20 Close