Invariant manifolds of nonautonomous models neural networks
M.L. Kolomiets1, A.N. Sakharov2
Annotation | We consider the problem of the Structure of invariant sets in nonautonomous systems describing the dynamics of neural networks. It is shown that normal hyperbolic manifolds of non-linear extensions of quasiperiodic flows on the torus is a bundle over the torus with a homogeneous space of compact matrix Lie group as a fiber. |
---|---|
Keywords | flows extensions on compact spaces, normally hyperbolic diversity, attractors, minimal sets, neural networks |
1Assistant professor of department of higher mathematics, Nizhny Novgorod State Agricultural Academy, Nizhny Novgorod; math@agri.sci-nnov.ru.
2Assistant professor of department of higher mathematic, Nizhny Novgorod State Agricultural Academy, Nizhny Novgorod; ansakharov2008@yandex.ru.
Citation: M.L. Kolomiets, A.N. Sakharov, "[Invariant manifolds of nonautonomous models neural networks]", Zhurnal Srednevolzhskogo matematicheskogo obshchestva,14:2 (2012) 74–80 (In Russian)