Connectionism
Jump to navigation
Jump to search
Connectionism
The approach to AI, and to modelling the mind more broadly, that builds intelligent behaviour out of networks of simple interconnected units adjusting their connections through learning, rather than explicit symbolic rules. It is the intellectual ancestor of the neural network and deep learning methods behind virtually all current AI systems, and stands in contrast to symbolic AI, the older, rule-based school it eventually displaced. (See also: Symbolic AI, Neural network)