Learning overhypotheses with hierarchical Bayesian models
Date
2007
Authors
Kemp, C.
Perfors, A.
Tenenbaum, J.
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Journal article
Citation
Developmental Science, 2007; 10(3):307-321
Statement of Responsibility
Charles Kemp, Amy Perfors and Joshua B. Tenenbaum
Conference Name
Abstract
Inductive learning is impossible without overhypotheses, or constraints on the hypotheses considered by the learner. Some of these overhypotheses must be innate, but we suggest that hierarchical Bayesian models can help to explain how the rest are acquired. To illustrate this claim, we develop models that acquire two kinds of overhypotheses – overhypotheses about feature variability (e.g. the shape bias in word learning) and overhypotheses about the grouping of categories into ontological kinds like objects and substances.