Staying afloat on Neurath's boat - Heuristics for sequential causal learning

Causal models are key to flexible and efficient exploitation of the environment. However, learning causal structure is hard, with massive spaces of possible models, hard-to-compute marginals and the need to integrate diverse evidence over many instances. We report on two experiments in which participants learnt about probabilistic causal systems involving three and four variables from sequences of interventions. Participants were broadly successful, albeit exhibiting sequential dependence and floundering under high background noise. We capture their behavior with a simple model, based on the “Neurath’s ship” metaphor for scientific progress, that neither maintains a probability distribution, nor computes exact likelihoods.

[1]  Illtyd Trethowan Causality , 1938 .

[2]  Willard Van Orman Quine,et al.  Word and Object , 1960 .

[3]  I. Lakatos Falsification and the Methodology of Scientific Research Programmes , 1976 .

[4]  Gregory F. Cooper,et al.  A Bayesian Method for the Induction of Probabilistic Networks from Data , 1992 .

[5]  Jun S. Liu,et al.  Sequential Monte Carlo methods for dynamic systems , 1997 .

[6]  P. Cheng From covariation to causation: A causal power theory. , 1997 .

[7]  Peter Dayan,et al.  Expected and Unexpected Uncertainty: ACh and NE in the Neocortex , 2002, NIPS.

[8]  Joshua B. Tenenbaum,et al.  Inferring causal networks from observations and interventions , 2003, Cogn. Sci..

[9]  S. Sloman,et al.  The advantage of timely intervention. , 2004, Journal of experimental psychology. Learning, memory, and cognition.

[10]  Stergios B. Fotopoulos,et al.  Introduction to Modern Nonparametric Statistics , 2004, Technometrics.

[11]  S. Sloman,et al.  Time as a guide to cause. , 2006, Journal of experimental psychology. Learning, memory, and cognition.

[12]  Aaron C. Courville,et al.  The rat as particle filter , 2007, NIPS.

[13]  A. Yuille,et al.  Bayesian generic priors for causal learning. , 2008, Psychological review.

[14]  Philip M. Fernbach,et al.  Causal learning with local computations. , 2009, Journal of experimental psychology. Learning, memory, and cognition.

[15]  D. Shanks,et al.  Learning in a changing environment. , 2010, Journal of experimental psychology. General.

[16]  Joshua B. Tenenbaum,et al.  Edge replacement and nonindependence in causation , 2010 .

[17]  Adam N Sanborn,et al.  Rational approximations to rational models: alternative algorithms for category learning. , 2010, Psychological review.

[18]  Ralf Mayrhofer,et al.  Heuristics in Covariation-based Induction of Causal Models: Sufficiency and Necessity Priors , 2011, CogSci.

[19]  Bob Rehder,et al.  Decisions to intervene on causal systems are adaptively selected , 2014, CogSci.

[20]  Thomas L. Griffiths,et al.  One and Done? Optimal Decisions From Very Few Samples , 2014, Cogn. Sci..

[21]  David A. Lagnado,et al.  The order of things: Inferring causal structure from temporal patterns , 2014, CogSci.