ABSTRACT
This chapter presents various Chaotic neural networks (NN) that do, indeed, possess the abovementioned Pattern recognition (PR) properties. The four best approaches for PR are template matching, statistical classification, syntactic or structural recognition, and artificial NNs (ANNs). The salient characterizing phenomenon that a chaotic PR system must exhibit is, on the one hand, chaos for untrained patterns. In the training phase, one would present the chaotic PR system with a set of patterns, and it thus “learns” the weights of the chaotic NNs (CNN). The Adachi neural network possesses rich dynamical phenomena such as associative memory (AM), and some PR and chaotic properties as one or more of its parameters change. The chapter discusses the AM and PR capabilities of the modified AdNN (M-AdNN) using the strategy. The trained M-AdNN again demonstrated a periodic response when the nonnoisy external stimuli were applied, after an initial nonperiodic transient phase.
