Part of Advances in Neural Information Processing Systems 10 (NIPS 1997)
Peter Dayan, Theresa Long
Conditioning experiments probe the ways that animals make pre(cid:173) dictions about rewards and punishments and use those predic(cid:173) tions to control their behavior. One standard model of condition(cid:173) ing paradigms which involve many conditioned stimuli suggests that individual predictions should be added together. Various key results show that this model fails in some circumstances, and mo(cid:173) tivate an alternative model, in which there is attentional selection between different available stimuli. The new model is a form of mixture of experts, has a close relationship with some other exist(cid:173) ing psychological suggestions, and is statistically well-founded.