The Rational Basis of Representativeness

By Josh Tenenbaum, T. Griffiths. 2001

How researchers see this paper

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

Citations in the corpus

Machine Teaching: An Inverse Problem to Machine Learning and an Approach Toward Optimal Education (2015, background): “Another example is a Bayesian learner with a prior conjugate to D, where A(D) is simply the posterior distribution (Zhu 2013; Tenenbaum and Griffiths 2001; Rafferty and Griffiths 2010).”
Machine Teaching for Bayesian Learners in the Exponential Family (2013, background): “Tenenbaum and Griffiths is the first to note that representative data is one that maximizes the posterior probability of the target model [22].”
A rational account of pedagogical reasoning: teaching by, and learning from, examples. (2014, unknown): “One distinction that has been introduced in the literature on concept learning is the difference between weak sampling and strong sampling (Hsu & Griffiths, 2009; Tenenbaum, 1999; Tenenbaum & Griffiths, 2001a).”
On the nature and origin of intuitive theories : learning, physics and psychology (2015, background): “This provides a form of implicit negative evidence [175], useful as an inductive bias when only positive facts of a domain are observed.”
Toward a general, scaleable framework for Bayesian teaching with applications to topic models (2016, background): “Zhu [10] offers a formalization for exponential family distributions [see also 7].”
Theory learning as stochastic search in the language of thought (2012, background): “…likelihood we see that theory T predicts data D well if it assigns high prior P ( M T ) to models M that make the data probable under the observation process P ( D M ). negative evidence (Tenenbaum & Grifﬁths, 2001), useful as an inductive bias when only positive facts of a domain are observed.”
Optimal Language Learning: The Importance of Starting Representative (2010, unknown): “Tenenbaum and Griffiths (2001) suggested that R(d,θ∗) be considered a measure of the “representativeness” of d relative to θ∗, being an indicator of the strength of evidence that d provides in favor of θ∗ relative to other values of θ.”