TY - GEN
T1 - A Nonparametric Bayesian Model of Multi-Level Category Learning
AU - Canini, Kevin R.
AU - Griffiths, Thomas L.
N1 - Publisher Copyright:
Copyright © 2011, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
PY - 2011/8/11
Y1 - 2011/8/11
N2 - Categories are often organized into hierarchical taxonomies, that is, tree structures where each node represents a labeled category, and a node's parent and children are, respectively, the category's supertype and subtypes. A natural question is whether it is possible to reconstruct category taxonomies in cases where we are not given explicit information about how categories are related to each other, but only a sample of observations of the members of each category. In this paper, we introduce a nonparametric Bayesian model of multi-level category learning, an extension of the hierarchical Dirichlet process (HDP) that we call the tree-HDP. We demonstrate the ability of the tree-HDP to reconstruct simulated datasets of artificial taxonomies, and show that it produces similar performance to human learners on a taxonomy inference task.
AB - Categories are often organized into hierarchical taxonomies, that is, tree structures where each node represents a labeled category, and a node's parent and children are, respectively, the category's supertype and subtypes. A natural question is whether it is possible to reconstruct category taxonomies in cases where we are not given explicit information about how categories are related to each other, but only a sample of observations of the members of each category. In this paper, we introduce a nonparametric Bayesian model of multi-level category learning, an extension of the hierarchical Dirichlet process (HDP) that we call the tree-HDP. We demonstrate the ability of the tree-HDP to reconstruct simulated datasets of artificial taxonomies, and show that it produces similar performance to human learners on a taxonomy inference task.
UR - http://www.scopus.com/inward/record.url?scp=84906868247&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84906868247&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:84906868247
T3 - Proceedings of the 25th AAAI Conference on Artificial Intelligence, AAAI 2011
SP - 307
EP - 312
BT - Proceedings of the 25th AAAI Conference on Artificial Intelligence, AAAI 2011
PB - AAAI press
T2 - 25th AAAI Conference on Artificial Intelligence, AAAI 2011
Y2 - 7 August 2011 through 11 August 2011
ER -