Parallel Computation of Fractal Sets with the Help of Neural Networks and Cellular Automata
by Vasily Severyanov
Abstract:
Fractal sets of a broad class are described by Deterministic Iterated Function Systems. It has been shown by J. Stark that one can build a binary asymmetric Neural Network whose attractor gives an approximation of the corresponding fractal set. The author of the article has suggested a version of the Neural Network Algorithm which is more convenient and efficient in some cases. Here we show a way of going from Deterministic Iterated Function Systems to a special class of Cellular Automata and give a hint how our Neural Network Algorithm can be converted to become a Cellular Automaton Algorithm. Cellular automata are simpler than neural networks and well suited for parallel implementation.
Keywords: theory
Source:
V. Severyanov, Parallel Computation of Fractal Sets with the Help of Neural Networks and Cellular Automata. In V. Malyshkin (ed.),
Parallel Computing Technologies: Proceedings of the 4th International Conference,
Lect. Notes in Comp. Sci., Vol. 1277, Springer, 1997, pp. 109-114
