Algorithms for Large Scale Markov Blanket Discovery

By Ioannis Tsamardinos, Constantin F. Aliferis, Alexander R. Statnikov. The Florida AI Research Society, 2003

632 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/0a332757b8a52c0aea548c1f9fde106d397e5334

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

A review of feature selection methods based on mutual information (2013, methodology): “Fast MB discovery algorithms have been developed for the case of distributions that are faithful to a Bayesian Network [58, 59].”
Algorithms for Discovery of Multiple Markov Boundaries (2013, methodology): “Finally, since very fast single Markov boundary induction algorithms have been recently introduced (Aliferis et al., 2010a, 2003a; Peña et al., 2007; Tsamardinos et al., 2003a,b), the overall TIE∗ operation is very fast.”
HITON: A Novel Markov Blanket Algorithm for Optimal Variable Selection (2003, methodology): “The widely-cited Koller-Sahami algorithm is unsound, cannot be run in datasets as large as the ones used in our experiments, and was recently shown to perform worse than other (sound) algorithms [26].”
Time and sample efficient discovery of Markov blankets and direct causal relations (2003, methodology): “MMMB is compared with all previously known algorithms for inducing Markov Blankets, namely the Incremental Association Markov Blanket (IAMB) algorithm [11], the Grow-Shrink (GS) algorithm [5], and the Koller-Sahami algorithm (KS) [4].”
A Survey of Learning Causality with Data (2018, background): “Hybrid algorithms [153, 159] that exploit principled ways to combine CB and SB algorithms have also attracted considerable attention.”
Toward Provably Correct Feature Selection in Arbitrary Domains (2009, methodology): “Othe r algorithms include the GS algorithm [10], originally developed for learning of the structure of a Bayesian network of a domain, and extensions to it [11] including the recent MMMB algorithm [12] .”
Scaling-Up Bayesian Network Learning to Thousands of Variables Using Local Learning Techniques (2003, methodology): “The Incremental Association Markov Blanket algorithm [12] improves over the Grow-Shrink by employing a dynamic variable ordering heuristic.”
Advances in variable selection methods I: Causal selection methods versus stepwise regression and principal component analysis on data of known and unknown functional relationships (2012, methodology): “…the causal methods include: Grow-Shrink, GS (Margaritis and Thrun, 1999); a variant of Incremental Association Markov Boundary (IAMB), interIAMBnPC (Tsamardinos et al., 2003); Local Causal Discovery, LCD2 (Cooper, 1997); HITON Markov Blanket, HITON-MB (Aliferis et al., 2003a); and First Order…”
Masked Gradient-Based Causal Structure Learning (2019, methodology): “In practice, heuristic methods such as Incremental Association Markov Blanket (IAMB) (Tsamardinos et al., 2003) can be used to search for the desired subset.”
Learning Bayesian Networks with the bnlearn R Package (2009, methodology): “• Interleaved Incremental Association ( inter.iamb ): another variant of IAMB which uses forward stepwise selection (Tsamardinos et al. 2003) to avoid false positives in the Markov blanket detection phase.”
Learning Bayesian networks: approaches and issues (2011, methodology): “Aliferis and Tsamardinos (2002) and Tsamardinos et al. (2003a, 2003b) focus on learning local features of a Bayesian network.”
Causal Feature Selection (2007, methodology): “MMMB [34] returns all members of the Markov Blanket of the target and some false positives, which may be eliminated with various techniques, if needed.”
Conditional Likelihood Maximisation: A Unifying Framework for Information Theoretic Feature Selection (2012, methodology): “A common heuristic approach is a sequential search considering features one-by-one for addition/removal ; this is used for example in Markov Blanket learning algorithms such as IAMB (Tsamardinos et al., 2003).”
Local Causal and Markov Blanket Induction for Causal Discovery and Feature Selection for Classification Part I: Algorithms and Empirical Evaluation (2010, methodology): “Several empirical studies have verified their applicability (Tsamardinos et al., 2003b; Spirtes et al., 2000; Glymour and Cooper, 1999; Aliferis and Cooper, 1994).”
The max-min hill-climbing Bayesian network structure learning algorithm (2006, methodology): “If 〈G, P 〉 and 〈G′, P 〉 are two faithful Bayesian networks (to the same distribution), then for any variable T , it is the case that PCGT = PC G′ T (Verma and Pearl, 1990; Pearl and Verma, 1991; Tsamardinos et al., 2003c).”
Information-Theoretic Inference of Large Transcriptional Regulatory Networks (2007, background): “Furthermore, MRNET beneﬁts from some additional properties which are common to all the feature selection strategies for network inference [26, 27], as follows.”