A class of computational models that consist of a number of interconnected simple processing units. Often the units are arranged in several layers. Connectionist models learn from experience, and have been used to explain several aspects of infant cognitive and behavioral development, as well as developmental disorders. In addition, connectionist models and neural networks have given rise to computational neuroscience, another example of a new emerging interdiscipline that draws on computer science, neuroscience and applied mathematics to gain sights into the functioning of the nervous system. The origins of computational neuroscience reside in the work Alan L. Hodgkin (1914-1998) and Andrew F. Huxley in deriving the first mathematical model of the action potential for which they received the Nobel Prize for Physiology or Medicine in 1963 (together with John C. Eccles, 1903-1997). As a term, connectionism was first mentioned by Edward Lee Thorndike (1874-1949) in 1898 with regard to the way in which cats learn by means of incremental stages.
See Action potential, Activation (in a connectionist model), Activity-dependent organization, Artificial intelligence, Auto-encoder networks, Backprogagation, Cognitive-functional approach, Computational models, Connectionism, Interdiscipline, Neural net, Neuroconstructivist theories, Non-linear associator, On-line emergence, Phonological pathway, Processing units