Dissipative Structures Emerging in an Abstract Representation of a Neural Network Satisfying Hebb’s Hypothesis
Giovanni Maria Guazzo · European Journal of Theoretical and Applied Sciences · 2025
This study presents a mathematical model of the learning process satisfying Hebb's assumptions: i.e. given any external input, a system of interconnected elements must necessarily give rise to an appropriate pattern of stable spatiotemporal activity. The author, adopting this point of view, did not use the models that consider learning as a modification of specific interconnection coefficients; instead, he made use of the continuous mathematical models of morphogenesis, identifying learning with the process of the formation of a dissipative structure; i.e., of a stable non-equilibrium structure, which is formed from a state of homogeneous equilibrium, as soon as certain parameters exceed specific critical values.