TENSOR PRODUCT NEURAL NETWORKS ANDAPPROXIMATION OF DYNAMICAL SYSTEMSAjit

@inproceedings{Dingankar1996TENSORPN,
  title={TENSOR PRODUCT NEURAL NETWORKS ANDAPPROXIMATION OF DYNAMICAL SYSTEMSAjit},
  author={T. Dingankar and Irwin W. Sandberg and A. U.S. and Ajit},
  year={1996}
}
We consider the problem of approximating any member of a large class of input-output operators of nonlinear dynam-ical systems. The systems need not be shift invariant, and the system inputs need not be continuous. We introduce a family of \tensor product" dynamical neural networks, and show that a certain continuity condition is necessary and suucient for the existence of arbitrarily good approximations using this family.