| Popis: |
Fuzzy cognitive maps are recurrent neural networks, applied for modelling and simulation of complex dynamic systems. They have been successfully applied to many engineering problems. The conclusion about the system depends on the mathematical behaviour of an iteration, namely, a first-order recursion. The first-order dynamics have some limitations since it takes into account only the previous time step. To overcome these limitations higher-order memory-based fuzzy cognitive maps have been introduced, which use a sequence of preceding states to determine the next one. In this paper, some dynamical properties of higher-order fuzzy cognitive maps are analyzed. Particularly, the existence and uniqueness of equilibrium points and the stability are discussed. |
