Complexity Symbiosis of Glia-Neuron Cells and Computational Cybernetics of Hopfield Recurrent Network: Novel Neuron Model

Science of artificial neural networks and relevant computing mechanisms since McCullock-Pitts artificial neuron (1943) up via neurons and networks of Anderson (1972), Barto et al (1983), Grosberg (1967, 1976), Hopfield (1982, 1984), Kohonen (1972) to Kasabov’s evolving connectionist systems with spiking-neurons (2003) have undergone developments beyond any conceivable predictions. The computational efficiency and functionality of all kinds of neural network implies stable operating steady-state equilibrium is fast established and guaranteed. In parallel, Neurophysiology has yielded many insights Gayton-Hall (2006) converging to paradigm of systems biology. It appeared, on the crossroad of these findings with Hilbert’s Thirteen problem and Kolmogorov’s Superposition Representations in conjunction with Lyapunov foundations of stability and LaSalle invariance principle certain delicate subtle issues emerged Siljak (2008) and Sprecher (2017). This re-thinking the foundations of neural networks via the quest for parallels between artificial and living neurons is believed to open a new horizon. This belief follows obtained results on cultured-neuron controllers and recurrent neural networks with time-varying delays. A closer look into how animal and/or human brain cells can be cultivated as a controlling brain for a mobile robot (physical body) such that can move around and interact with the world. In turn, a new kind of artificial intelligence may be created, which is emulated by stabilized complex highly non-linear complex neural network system.