Almost Periodic Solutions in Distribution Sense for Quaternion-Valued Stochastic Delayed Neural Networks

In this paper, we consider quaternion-valued stochastic delayed neural networks. We first obtain the existence of almost periodic solutions in distribution sense by employing the contraction mapping principle. Then by using stochastic analysis and inequality techniques, we obtain the mean square global exponential stability of the almost periodic solutions of the considered neural networks. Finally, we present a numerical example to illustrate the feasibility of our results.