Development of a stimulus-response learning model as a fundamental data processing element
Brady, Charles Darnell
Silbert, E. E.
Master of Science
A stimulus-response learning model was developed and tested in a simulated paired-associate learning experiment. The behavior of the model was compared with characteristic human behavior. The model was sufficiently successful to give weight to the stimulus-response approach to learning. Although it was revealed that paired-associate learning involves complexities which cannot be explained by a single simple stimulus-response learning model, it was also demonstrated that these devices or stimulus-response pair association units (PAU's) are capable of complex data processing when properly interconnected. It was shown that a Turing machine as well as a general purpose digital computer can be constructed entirely of PAU's thus demonstrating their capability. Finally, a speculative model of the human thought processes is included which consists only of pair-association units.