This chapter shows how an attractor network can model probabilistic decision making. For decision making, the attractor network is trained to have two or more attractor states, each of which corresponds to one of the decisions. Each attractor set of neurons receives a biasing input which corresponds to the evidence in favour of that decision. The model not only shows how probabilistic decision making could be implemented in the brain, but also how the evidence can be accumulated over long periods of time because of the integrating action of the attractor's short-term memory network.
Oxford Scholarship Online requires a subscription or purchase to access the full text of books within the service. Public users can however freely search the site and view the abstracts and keywords for each book and chapter.
If you think you should have access to this title, please contact your librarian.