This paper aims to theoretically prove by applying Marotto's Theorem t
hat both transiently chaotic neural networks (TCNN) and discrete-time
recurrent neural networks (DRNN) have chaotic structure. A significant
property TCNN and DRNN is that they have only one bounded fixed point
, when absolute values of the self-feedback connection weights in TCNN
and the difference time in DRNN are sufficiently large. We show that
this unique fixed point tan actually evolve into a snap-back repeller
which generates chaotic structure, if several conditions are satisfied
. On the other hand, by using the Lyapunov functions, we also derive;
sufficient conditions on asymptotical stability for symmetrical versio
ns of both TCNN and DRNN, under which TCNN and DRNN asymptotically con
verge to a fixed point. Furthermore, related bifurcations are also con
sidered in this paper. Since both TCNN and DRNN are not special but si
mple and general, the obtained theoretical results hold for a wide cla
ss of discrete-time neural networks. To demonstrate the theoretical re
sults of this paper better, several numerical simulations ale provided
as illustrating examples.