Being Bayesian About Network Structure. A Bayesian Approach to Structure Discovery in Bayesian Networks

By Nir Friedman, Daphne Koller. 2004

985 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Statistics and Probability, Probabilistic & Bayesian Methods

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 Statistics and Probability

In Probabilistic & Bayesian Methods

Citations in the corpus

Causal Inference through a Witness Protection Program (2014, neutral): “An approach analogous to (Friedman and Koller, 2003) is necessary, which we leave as future work.”
Venture: a higher-order probabilistic programming platform with programmable inference (2014, background): “, 2007; Rasmussen and Williams, 2006; Griffiths and Ghahramani, 2005), probabilistic models over worlds with a priori unknown numbers of objects, models for learning the structure of graphical models (Heckerman, 1998; Friedman and Koller, 2003; Mansinghka et al., 2006), models for inductive learning”
DAG-GNN: DAG Structure Learning with Graph Neural Networks (2019, background): “Many bootstrap (Friedman et al., 1999) and sampling-based structure learning algorithms (Madigan et al., 1995; Friedman & Koller, 2003; Eaton & Murphy, 2012; Grzegorczyk & Husmeier, 2008; Niinim¨aki & Koivisto, 2013; Niinimaki et al., 2012; He et al., 2016) are also proposed to tackle the expensive…”
UvA-DARE ( Digital Academic Repository ) Causal Discovery with Continuous Additive Noise Models (2014, positive): “In this sense the set of true orderings This result is useful, for example, if the search over structures is performed in the space of permutations rather than in the space of DAGs (e.g. Friedman and Koller, 2003; Teyssier and Koller, 2005; B¨uhlmann et al., 2013).”
Advanced Lectures on Machine Learning (2004, methodology): “An alternative to trying to find the most probable graph are methods that sample over the posterior distribution of graphs [58].”
Structure Discovery in Bayesian Networks by Sampling Partial Orders (2016, methodology): “Friedman and Koller (2003) obtained a significantly faster mixing chain by operating, not directly on DAGs, but in the much smaller and smoother space of node orderings, or linear orders on the nodes more formally.”
Bayesian Model Averaging: A Systematic Review and Conceptual Classification (2015, neutral): “…model priors by the expected behavior of the cell cultures under study, the analytical considerations of the Bayesian networks of interest in Friedman and Koller (2003), the combination of multiple stakeholders in Mäntyniemi et al. (2013) and the proposal of a cross-model correlations to…”
Unsupervised Learning (2018, neutral): “An alternative to trying to find the most probable graph are methods that sample over the posterior distribution of graphs [58].”
Learning Bayesian networks: approaches and issues (2011, methodology): “A further improvement on this general scheme is given by the system of Grzegorczyk and Husmeier (2008) that improves the convergence of the classical approach and that of Liang and Zhang (2009) who use a stochastic approximation Monte Carlo (SAMC) method. Riggelsen and Feelders (2005) extend MC(3) t”
D’ya Like DAGs? A Survey on Structure Learning and Causal Discovery (2021, background): “[252] 2020 score Granger no Bi Order-MCMC [61] 2003 sampling yes yes yes no DAG OG [54] 2008 sampling yes yes yes yes DAG EE-DAG [269] 2011 sampling yes yes yes yes DAG ZIPBN [35] 2020 sampling yes no yes no DAG LiNGAM [216] 2006 asymmetries yes no yes no DAG LV LiNGAM [103] 2008 asymmetries no yes ”
Pattern Recognition and Machine Learning (2006, neutral): “However, there is also interest in going beyond the inference problem and learning the graph structure itself from data (Friedman and Koller, 2003).”
BCD Nets: Scalable Variational Approaches for Bayesian Causal Discovery (2021, background): “For causal discovery, naive Monte-Carlo sampling or enumeration of the possible DAGs quickly becomes intractable, since the number of possible DAGs over a model grows superexponentially with the number of variables [18].”
Network Inference and Biological Dynamics (2011, methodology): “Otherwise, Markov chain Monte Carlo (MCMC) methods can be used to explore an eﬀectively smaller model space [Ellis and Wong (2008); Friedman and Koller (2003)].”
Supervised Machine Learning: A Review of Classification Techniques (2007, positive): “Friedman & Koller (2003) proposed a new approach for this task and showed how to efficiently compute a sum over the exponential number of networks that are consistent with a fixed order over networks.”