Computer modeling of learning processes

Learning can be said to take place when an entity changes its behavior as a result of past experience. Many attempts have been made at the computer modeling of the learning processes in part or in whole. This paper focuses on those attempts and the theories that were their basis.Mechanisms needed for various functions of learning are compared and contrasted. Several theories are discussed, including learning machines and perceptrons, general pattern recognition, paired associate and serial rote learning and operant and classical conditioning. Also discussed are attempts to program induction and discovery into learning models.Programs are also presented illustrating some of these attempts to model learning processes.