THE GENERALIZATION FUNCTION IN THE PROBABILITY LEARNING EXPERIMENT.
Abstract
In this report, the author formulates and studies some methods for obtaining generalization functions from learning data. First he considers mathematical questions and concludes that the generalization function defined with respect to a slight modification of a familiar learning model is essentially determined by the behavior of the individual subject in one experiment. Next it is shown that generalization functions obtained by application of the methods can be used to predict certain empirical functions with great accuracy. Finally, he studies the empirical generalization functions and attempts to describe and account for the relationship between the function and distribution of reinforcements.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 03, 1965
- Accession Number
- AD0622587
Entities
People
- Michael V. Levine
Organizations
- Stanford University