Fixed points and ppth moment exponential stability of stochastic delayed recurrent neural networks with impulses

New sufficient conditions for ppth moment exponential stability of a class of impulsive stochastic delayed recurrent neural networks are presented by using fixed point theory. Our results neither require the boundedness, monotonicity and differentiability of the activation functions nor differentiability of the time varying delays. A class of impulsive delayed neural networks without stochastic perturbations are also considered. An example is given to illustrate our main results.