Stochastic learning theory applied to choice experiments with rats, dogs, and men

Abstract

The usual dilemma faced in selecting mathematical models for use in describing and predicting dynamic behavioral choice experiments is that a stochastic model that seems complex enough to prove valid will very likely turn out to require solution of impossibly difficult statistical estimation problems. Here, this dilemma has been resolved by taking advantage of the capability of modern electronic computers to solve very complex numerical problems by trial and error, so that just one computational technique becomes adequate for all numerical estimation problems. A variety of stochastic behavioral models—old and new, simple and complex—are used successfully to analyze data from a variety of choice experiments—old and new, simple and complex—confirming or revising earlier work of some other authors and here also extending both the experimental and the analytical findings.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 10 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.