Title :
Forward accessibility for recurrent neural networks
Author :
Albertini, Francesca ; Pra, Paolo Dai
Author_Institution :
Dipartimento di Matematica Pura e Applicata, Padova Univ., Italy
fDate :
11/1/1995 12:00:00 AM
Abstract :
Gives an algebraic characterization of forward accessibility for recurrent neural networks
Keywords :
algebra; discrete time systems; recurrent neural nets; algebraic characterization; forward accessibility; recurrent neural networks; Control systems; Controllability; Differential equations; Linear systems; Neural networks; Nonlinear control systems; Observability; Q measurement; Recurrent neural networks; State-space methods;
Journal_Title :
Automatic Control, IEEE Transactions on
