Efficient Markov Network Structure Discovery Using Independence Tests
From MaRDI portal
Abstract: We present two algorithms for learning the structure of a Markov network from data: GSMN* and GSIMN. Both algorithms use statistical independence tests to infer the structure by successively constraining the set of structures consistent with the results of these tests. Until very recently, algorithms for structure learning were based on maximum likelihood estimation, which has been proved to be NP-hard for Markov networks due to the difficulty of estimating the parameters of the network, needed for the computation of the data likelihood. The independence-based approach does not require the computation of the likelihood, and thus both GSMN* and GSIMN can compute the structure efficiently (as shown in our experiments). GSMN* is an adaptation of the Grow-Shrink algorithm of Margaritis and Thrun for learning the structure of Bayesian networks. GSIMN extends GSMN* by additionally exploiting Pearls well-known properties of the conditional independence relation to infer novel independences from known ones, thus avoiding the performance of statistical tests to estimate them. To accomplish this efficiently GSIMN uses the Triangle theorem, also introduced in this work, which is a simplified version of the set of Markov axioms. Experimental comparisons on artificial and real-world data sets show GSIMN can yield significant savings with respect to GSMN*, while generating a Markov network with comparable or in some cases improved quality. We also compare GSIMN to a forward-chaining implementation, called GSIMN-FCH, that produces all possible conditional independences resulting from repeatedly applying Pearls theorems on the known conditional independence tests. The results of this comparison show that GSIMN, by the sole use of the Triangle theorem, is nearly optimal in terms of the set of independences tests that it infers.
Recommendations
- Efficient identification of independence networks using mutual information
- Hierarchical models for independence structures of networks
- A simple method for testing independencies in Bayesian networks
- A note on testing conditional independence for social network analysis
- On a simple method for testing independencies in Bayesian networks
- Bayesian network structure learning: hybridizing complete search with independence tests
- Mutual conditional independence and its applications to model selection in Markov networks
- Identifying independence in bayesian networks
Cited in
(12)- Probabilistic graphical models and Markov networks
- Objective Bayesian Nets for Integrating Consistent Datasets
- A Gaussian noise model based algorithm to construct Markov networks
- AMP chain graphs: minimal separators and structure learning algorithms
- scientific article; zbMATH DE number 7370579 (Why is no real title available?)
- A note on testing conditional independence for social network analysis
- Sparse model selection in the highly under-sampled regime
- Mutual conditional independence and its applications to model selection in Markov networks
- Improving Markov network structure learning using decision trees
- Blankets joint posterior score for learning Markov network structures
- Learning chordal Markov networks via stochastic local search
- The IBMAP approach for Markov network structure learning
This page was built for publication: Efficient Markov Network Structure Discovery Using Independence Tests
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3651469)
