“…A BAMNN behaves as a two-layer pattern-matched hetero-associative content addressable memory, storing and recalling the pattern pairs [1]. Applications of BAMNN in physical systems include pattern recognition, signal and image process, and automatic control engineering [2,3]. In recent years, artificial neural networks, such as BAMNN, cellular neural networks, Cohen-Grossberg 452 R.-S. GAU, J.-G. HSIEH AND C.-H. LIEN neural networks, and Hopfield neural networks, have been developed to solve various problems and to research in some topics .…”