Influences of Perturbation of Training Pattern Pairs on Morphological Bidirectional Associative Memories

An important issue in design and implementation a neural network is that perturbations of training pattern pairs may cause some disadvantages to outputs. How the perturbations of training pattern pairs in Morphological Bidirectional Associative Memories (MBAMs) influence on the outputs is discussed in this paper. We define the outputs' max error to evaluate the robustness of the MBAMs. The related theorem and example show that the MBAMs have good robustness on the perturbation of training pattern pairs. This is of important sense in MBAMs' practical applications, such as analysis of the networks' performances, selection or establishment of the learning algorithms and acquisition the training pattern pairs.