Weight decay is a common technique for improving generalization in the field of neural networks. If weight decay is added to the Rescorla-Wagner model, a desirable result emerges in that the effects of prior learning are not expected to be permanent. This revised model accommodates an experimental finding called unblocking that was claimed to refute the Rescorla-Wagner model.
Get full access to this article
View all access options for this article.
References
1.
AzorlosaJ. L.CicalaG. A. (1988) Increased conditioning in rats to a blocked CS after the first compound trial. Bulletin of the Psychonomic Society, 26, 254–257.
2.
HintonG. E. (1986) Learning distributed representations of concepts. In Proceedings of the Eighth Annual Conference of the Cognitive Science Society. Mahwah, NJ: Erlbaum. Pp. 1–12.
3.
KaminL. J. (1969) Predictability, surprise, attention, and conditioning. In CambellB. A.ChurchR. M. (Eds.), Punishment and aversive behavior. New York: Appleton-Century-Crofts. Pp. 279–296.
4.
MillerR. R.BarnetR. C.GrahameN. J. (1995) Assessment of the Rescorla-Wagner model. Psychological Bulletin, 117, 363–386.
5.
RescorlaR. A.WagnerA. R. (1972) A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and nonreinforcement. In BlackA. H.ProkasyW. F. (Eds.), Classical conditioning: II. Current research and theory. New York: Appleton-Century-Crofts. Pp. 64–99.
6.
RumelhartD. E.DurbinR.GoldenR.ChauvinY. (1996) Back propagation: The basic theory. In SmolenskyP.MozerM. C.RumelhartD. E. (Eds.), Mathematical perspectives on neural networks. Mahwah, NJ: Erlbaum. Pp. 533–566.
7.
SuttonR. S.BartoA. G. (1981) Toward a modern theory of adaptive networks: Expectation and prediction. Psychological Review, 88, 135–170.
8.
YamaguchiM. (1999) New methods for solving the Rescorla-Wagner model. Behavioral Research Methods, Instruments, & Computers, 31, 684–688.
