Learning Bayesian networks from data: An information-theory based approach

By Jie Cheng, R. Greiner, Jonathan Kelly, David A. Bell, Weiru Liu. Artificial Intelligence, 2002

833 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/34e748341dd71f289d158696f3020e271458d2f6

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 Management Science and Operations Research

Citations in the corpus

Feature deduction and ensemble design of intrusion detection systems (2005, background): “The Markov blanket of a node n is the union of n's parents, n's children and the parents of n's children (Cheng et al., 2002).”
Supervised Machine Learning Algorithms: Classification and Comparison (2017, background): “A problem of BN classifiers is that they are not suitable for datasets with many features [4].”
A Scoring Function for Learning Bayesian Networks based on Mutual Information and Conditional Independence Tests (2006, methodology): “The associated statistical test has also been used by several learning algorithms based on conditional independence tests (Acid and de Campos, 2001; Cheng et al., 2002; de Campos and Huete, 2000; Spirtes et al., 1993).”
A kernel-based causal learning algorithm (2007, methodology): “well-known constraint-based PC algorithm, BN-PC (an information-theory-based refinement of IC) by Cheng et al. (2002) .”
Machine learning: a review of classification and combining techniques (2006, background): “A problem of BN classifiers is that they are not suitable for datasets with many features ( Cheng et al. 2002 ).”
Information-Theoretic Inference of Large Transcriptional Regulatory Networks (2007, methodology): “The adoption of mutual information in probabilistic model design can be traced back to Chow-Liu tree algo-rithm [3] and its extensions proposed by [7, 8].”
D’ya Like DAGs? A Survey on Structure Learning and Causal Discovery (2021, neutral): “PC [228] 1993 constraint yes yes yes no CPDAG FCI [229, 275] 1995 constraint no yes yes no POIPG CCD [200] 1996 constraint yes yes both no PAG TPDA [31] 2002 constraint yes yes yes no CPDAG CPC [193] 2006 constraint yes relaxed yes no CPDAG KCL [233] 2007 constraint yes yes yes no CPDAG ION [241] 20”
Local Causal and Markov Blanket Induction for Causal Discovery and Feature Selection for Classification Part I: Algorithms and Empirical Evaluation (2010, methodology): “…2002), and (iii) algorithm TPDA which restricts the class of distributions so that learning becomes from worst-case intractable to solvable in 4th degree polynomial time to the number of variables (and quadratic if prior knowledge about the ordering of variables is known) (Cheng et al., 2002a).”
Supervised Machine Learning: A Review of Classification Techniques (2007, neutral): “A problem of BN classifiers is that they are not suitable for datasets with many features (Cheng et al., 2002).”
Large engineering project risk management using a Bayesian belief network (2009, background): “A Bayesian belief network, also called a causal network or belief network, is a powerful tool for knowledge representation and reasoning under conditions of uncertainty (Cheng et al., 2002), and visually presents the probabilistic relationships among a set of variables (Heckerman, 1997).”
minet: A R/Bioconductor Package for Inferring Large Transcriptional Networks Using Mutual Information (2008, background): “The adoption of mutual information in network inference tasks can be traced back to the Chow and Liu's tree algo-rithm [13,14].”
Scaling-Up Bayesian Network Learning to Thousands of Variables Using Local Learning Techniques (2003, methodology): “For example, the public available versions of the PC [10] and the TPDA (also called PowerConstructor) [6] algorithms accept datasets with only 100 and 255 variables respectively, indicating the expectations of the inventors regarding their scalability.”
A survey of Bayesian Network structure learning (2021, methodology): “Figure 8, based on an example from Cheng et al. (2002), illustrates a subgraph of a true network to demonstrate the basic principles behind TPDA.”
Bayesian Networks and Decision Graphs (2001, methodology): “Improved algorithms, which theoretically should be more robust in face of flawed independence tests, are given by Margaritis and Thrun (1999) and Cheng et al. (2002). A discussion on observing causality can be found in (Pearl, 2000).”
Caveats for Causal Reasoning with Equilibrium Models (2001, methodology): “1 Constraint-Based Learning Constraint-Based (CB) learning methods [Verma and Pearl, 1991; Spirtes and Glymour, 1991; Pearl and Verma, 1991; Spirtes et al., 2000; Cheng et al., 2002] (also known as “conditional independence search” methods) build a CI model by systematically checking the data for co”
Learning Bayesian networks: approaches and issues (2011, methodology): “Another interesting approach is that of Cheng et al. (2002) and their three-phase dependency analysis (TPDA), who describe how an n variable DAG can be learned in O(n4) CI tests.”
Time and sample efficient discovery of Markov blankets and direct causal relations (2003, background): “This is the main idea behind constraint-based BN learning approaches [8, 3, 5].”
The max-min hill-climbing Bayesian network structure learning algorithm (2006, methodology): “The Three Phase Dependency Analysis (TPDA) algorithm (Cheng et al., 2002) follows the constraint-based paradigm and introduces the idea of using an information theoretic metric to search and test for conditional independencies.”