Abstract

We propose a decision theoretic approach for deciding which interventions to perform so as to learn the causal structure of a model as quickly as possible. Without such interventions, it is impossible to distinguish between Markov equivalent models, even given infinite data. We perform online MCMC to estimate the posterior over graph structures, and use importance sampling to find the best action to perform at each step. We assume the data is discrete-valued and fully observed.

Citations

1081	Local Computations with Probabilities on graphical structures and their applications to expert systems (with discussion - Lauritzen, Spiegelhalter
877	A Bayesian method for the induction of probabilistic networks from data - Cooper, Herskovits - 1992
766	Causality: Models, Reasoning, and Inference - Pearl
752	Learning Bayesian networks: The combination of knowledge and statistical data - Heckerman, Geiger, et al. - 1995
710	A tutorial on learning with Bayesian networks - Heckerman - 1995
663	Sequential Monte Carlo methods in practice. Statistics for engineering and information science - Doucet, Freitas, et al. - 2001
582	Queries and concept learning - Angluin - 1988
402	Active learning with statistical models - Cohn, Ghahramani, et al. - 1996
256	Selective sampling using the query by committee algorithm, Machine Learning 28 - Freund, Seung, et al. - 1997
236	Information-Based Objective Functions for Active Data - MacKay - 1992
213	Operations for learning with graphical models - Buntine - 1994
179	Equivalence and synthesis of causal models - Verma, Pearl
177	Sequential updating of conditional probabilities on directed graphical structures - Spiegelhalter, Lauritzen - 1990
168	Bayesian graphical models for discrete data - Madigan, York - 1995
161	Russel S: Learning the structure of dynamic probabilistic networks - Friedman, Murphy
156	Being Bayesian about network structure. a Bayesian approach to structure discovery in Bayesian networks - Friedman, Koller
111	Bayesian experimental design: A review - Chaloner, Verdinelli - 1995
71	Hoeffding races: Accelerating model selection search for classification and function approximation - Maron, Moore - 1993
63	Discovery of regulatory interactions through perturbation: Inference and experimental design - Ideker, Thorsson, et al. - 2000
53	Expected information as expected utility - Bernardo - 1979
52	A Bayesian approach to learning causal networks - Heckerman
51	Identification of gene regulatory networks by strategic gene disruptions and gene overexpressions - Akutsu, Kuahara, et al. - 1998
51	Causal discovery from a mixture of experimental and observational data - Cooper, Yoo - 1999
38	Improving Markov chain Monte Carlo model search for data mining. Machine learning 50 - Giudici, Castelo - 2003
38	D: Active Learning for Structure in Bayesian Networks - Tong, Koller
37	Sequential update of Bayesian network structure - Friedman, Goldszmidt - 1997
37	Troubleshooting under uncertainty - Heckerman, Breese - 1994
35	Bayesian model averaging and model selection for Markov equivalence classes of acyclic digraphs,” Comm in Statistics: Theory and Methods - Madigan, Andersson, et al. - 1996
21	An overview of the representation and discovery of causal relationships using Bayesian networks - Cooper - 1999
10	Learning boolean read-once formulas over generalized bases - Bshouty, Hancock, et al. - 1995
3	Insua. Decision analysis by augmented probability simulation - Bielza, Mueller, et al. - 1999
3	Learning from examples and membership queries with structured determinations - Tadepalli, Russell - 1998
2	Learning Bayesian networks: a uni for discrete and Gaussian domains - Heckerman, Geiger - 1995
1	Sampling methods for action selection in in diagrams - Ortiz, Kaelbling - 2000