Nonlinear forecasting and iterated function systems.
The theory of dynamical forecasting can be extended to iterated function systems. An unordered set of iterates may be sufficient to construct a simulation of the unknown dynamics. The underlying dynamical system may be nondeterministic: A random element may be allowed in the dynamics, and retrieved by our proposed procedures. A first application of these extensions pertains to the reconstruction of fractal attractors, and allows us to solve the so-called inverse problem in Iterated Functions Systems.
Journal Article↗