High-dimensional consistency in score-based and hybrid structure learning
From MaRDI portal
Publication:1991699
Abstract: Main approaches for learning Bayesian networks can be classified as constraint-based, score-based or hybrid methods. Although high-dimensional consistency results are available for constraint-based methods like the PC algorithm, such results have not been proved for score-based or hybrid methods, and most of the hybrid methods have not even shown to be consistent in the classical setting where the number of variables remains fixed and the sample size tends to infinity. In this paper, we show that consistency of hybrid methods based on greedy equivalence search (GES) can be achieved in the classical setting with adaptive restrictions on the search space that depend on the current state of the algorithm. Moreover, we prove consistency of GES and adaptively restricted GES (ARGES) in several sparse high-dimensional settings. ARGES scales well to sparse graphs with thousands of variables and our simulation study indicates that both GES and ARGES generally outperform the PC algorithm.
Recommendations
- Who learns better Bayesian network structures: accuracy and speed of structure learning algorithms
- Scaling up the greedy equivalence search algorithm by constraining the search space of equivalence classes
- Scaling up the greedy equivalence search algorithm by constraining the search space of equivalence classes
- Concave penalized estimation of sparse Gaussian Bayesian networks
- Estimating high-dimensional directed acyclic graphs with the PC-algorithm
Cites work
- 10.1162/153244302760200696
- 10.1162/153244303321897717
- A characterization of Markov equivalence classes for acyclic digraphs
- A constrained \(\ell _{1}\) minimization approach to sparse precision matrix estimation
- A formal approach to using data distributions for building causal polytree structures
- Approximating discrete probability distributions with dependence trees
- Causality. Models, reasoning, and inference
- Causation, prediction, and search
- Characterization and greedy learning of interventional Markov equivalence classes of directed acyclic graphs
- Estimating high-dimensional directed acyclic graphs with the PC-algorithm
- Estimating high-dimensional intervention effects from observational data
- Estimating the effect of joint interventions from observational data in sparse high-dimensional settings
- Extended Bayesian information criteria for model selection with large model spaces
- Geometry of the faithfulness assumption in causal inference
- High-dimensional Gaussian graphical model selection: walk summability and local separation criterion
- High-dimensional covariance estimation by minimizing \(\ell _{1}\)-penalized log-determinant divergence
- High-dimensional graphs and variable selection with the Lasso
- High-dimensional semiparametric Gaussian copula graphical models
- Independency relationships and learning algorithms for singly connected networks
- Model selection through sparse maximum likelihood estimation for multivariate Gaussian or binary data
- Order-independent constraint-based causal structure learning
- PC algorithm for nonparanormal graphical models
- Probabilistic graphical models.
- Scaling up the greedy equivalence search algorithm by constraining the search space of equivalence classes
- Sparse inverse covariance estimation with the graphical lasso
- Stability Selection
- The Adaptive Lasso and Its Oracle Properties
- The huge Package for High-dimensional Undirected Graph Estimation in R
- The max-min hill-climbing Bayesian network structure learning algorithm
- \(\ell_{0}\)-penalized maximum likelihood for sparse directed acyclic graphs
- \(\mathsf{PenPC}\): a two-step approach to estimate the skeletons of high-dimensional directed acyclic graphs
Cited in
(25)- Causal inference in partially linear structural equation models
- Bayesian Multi-Task Variable Selection with an Application to Differential DAG Analysis
- A survey on causal discovery: theory and practice
- Reconstruction of a directed acyclic graph with intervention
- scientific article; zbMATH DE number 7626781 (Why is no real title available?)
- Complexity analysis of Bayesian learning of high-dimensional DAG models and their equivalence classes
- Objective Bayesian Nets for Integrating Consistent Datasets
- Being Bayesian about learning Bayesian networks from ordinal data
- Partitioned hybrid learning of Bayesian network structures
- High-dimensional joint estimation of multiple directed Gaussian graphical models
- The reduced PC-algorithm: improved causal structure learning in large random networks
- Learning Bayesian network structures using weakest mutual-information-first strategy
- Estimation of joint directed acyclic graphs with Lasso family for gene networks
- scientific article; zbMATH DE number 7370576 (Why is no real title available?)
- Robust Causal Structure Learning with Some Hidden Variables
- Learning Moral Graphs in Construction of High-Dimensional Bayesian Networks for Mixed Data
- Causal discoveries for high dimensional mixed data
- Scoring matching approach: Learning high order Markov random fields
- Causal structure learning: a combinatorial perspective
- Who learns better Bayesian network structures: accuracy and speed of structure learning algorithms
- Scaling up the greedy equivalence search algorithm by constraining the search space of equivalence classes
- Efficient Learning of Quadratic Variance Function Directed Acyclic Graphs via Topological Layers
- Estimation of Gaussian directed acyclic graphs using partial ordering information with applications to DREAM3 networks and dairy cattle data
- Efficient Sampling and Structure Learning of Bayesian Networks
- Complete graphical characterization and construction of adjustment sets in Markov equivalence classes of ancestral graphs
This page was built for publication: High-dimensional consistency in score-based and hybrid structure learning
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1991699)
