MATHEMATICAL MODEL OF FRACTAL STRUCTURES RECOGNITION USING NEURAL NETWORK TECHNOLOGY

The article goes about the methods of training a neural network to recognize fractal structures with rotation iteration elements by means an improved randomized system functions. Parameters are used calculate complex parameters physical phenomena. They effective tool in scientific works and quantitative indicators technical tasks. calculation these is very difficult mathematical problem. This caused fact that it describe model image, determine iterative learning will allow you quickly first iterations based on finished image basing them functions (SRIF) process develop software for generating possibility rotating iterations. In its turn, this make possible form array data network. trained be able figures which build It help reproduce structure qualitatively. approach can three-dimensional structures. After setting fractal, geometric basis structure. future, may included recognizing objects under structures, example, masking nets.