UNIQUENESS OF WEIGHTS FOR RECURRENT NETS
Francesca Albertini, Eduardo D. Sontag, ∈ IRn×m · 2006
We study recurrent neural networks evolving either in discrete or in continuous time. The dynamics of such systems can be described by a set of either difference or differential equations of the following type (to simplify notations, we use the superscripts “+ ” and “. ” to denote time-shift and time-derivative respectively, and we omit the time arguments