Split quaternion-valued twin-multistate Hopfield neural networks (Q2173660)

Complex-valued Hopfield neural networks with a multistate activation function can be designed to process multi-level data for different kind of applications. In the present article the author concentrates on a special architecture, the so-called split quaternion-valued Hopfield neural networks with a T-multistate activation function which is understood as a pair of complex-valued multistate activation functions. The goal is to evaluate the performance of the network by means of simulations. There is a discussion on the obtained results.