A Rational Analysis of Rule-Based Concept Learning

By Noah D. Goodman, Josh Tenenbaum, J. Feldman, T. Griffiths. 2008

448 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Cultural Studies, Developmental and Educational Psychology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Cultural Studies

In Developmental and Educational Psychology

Citations in the corpus

Structure Learning in a Sensorimotor Association Task (2010, methodology): “This Bayesian account is entirely compatible with other rule-based approaches to concept learning [37], but a Bayesian estimator has to maintain a probability distribution over all alternatives at all times.”
One and Done? Optimal Decisions From Very Few Samples (2014, background): “Goodman et al. (2008) studied performance in classic categorization tasks, in which participants learn to discriminate two categories (A and B) by studying exemplars of each category, and are then asked to generalize the learned rules by categorizing new transfer items.”
Learning Programs: A Hierarchical Bayesian Approach (2010, methodology): “Bayesian inference has been used to induce logical formulae using a PCFG in several contexts, e.g., in representing natural language semantics (Piantadosi et al., 2008) and cognitive concepts (Goodman et al., 2008b).”
Theory Acquisition as Stochastic Search (2010, unknown): “The prior p(T |UT ) is Figure 3: The list of templates available to in the PHCG. the product of the probabilities of choices made at each point in this derivation.”
A rational account of pedagogical reasoning: teaching by, and learning from, examples. (2014, unknown): “Even Bayesian approaches, though required by the formalism to specify how the data are generated, have almost exclusively considered data to be the consequence of some random process (Anderson, 1991; Fried & Holyoak, 1984; Goodman et al., 2008; Gopnik et al., 2004; Griffiths & Tenenbaum, 2005).”
On the nature and origin of intuitive theories : learning, physics and psychology (2015, methodology): “This grammar and the Monte Carlo algorithms we use to sample or search over the theory posterior P (T |D,U) are based heavily on [61], which introduced the approach for learning single rule-based concepts rather than the larger law-based theory structures”
Theory of mind as inverse reinforcement learning (2019, background): “In addition, recent work suggests that representations originally developed to explain how people build complex concepts by composing simpler ones [14,66] may be useful.”