Instruments and Systems: Monitoring, Control, and Diagnostics Annotation << Back Development and implementation of information processing evolutionary algorithms D.P .BOGOLYUBOV, A.A. MIZIKIN, A.A. CHANKIN In this article we analyze actual problems in the area of in-formation processing, as well as potentially solving them with evolutionary algorithms developed by us. Basic principles of designing interval neural networks, learning algorithm for inter¬val neural network, genetic algorithm for determining suboptim¬al structure of neural networks and their CUDA-based software implementation are described. Key words: interval arithmetic, genetic algorithm, neural networks, evolutionary computations, GPGPU, CUDA. E-mail: bogolub@mail.ru