DAG-GNN: DAG Structure Learning with Graph Neural Networks

Abstract

Learning a faithful directed acyclic graph (DAG) from samples of a joint distribution is a challenging combinatorial problem, owing to the intractable search space superexponential in the number of graph nodes. A recent breakthrough formulates the problem as a continuous optimization with a structural constraint that ensures acyclicity (Zheng et al., 2018). The authors apply the approach to the linear structural equation model (SEM) and the least-squares loss function that are statistically well justified but nevertheless limited. Motivated by the widespread success of deep learning that is capable of capturing complex nonlinear mappings, in this work we propose a deep generative model and apply a variant of the structural constraint to learn the DAG. At the heart of the generative model is a variational autoencoder parameterized by a novel graph neural network architecture, which we coin DAG-GNN. In addition to the richer capacity, an advantage of the proposed model is that it naturally handles discrete variables as well as vector-valued ones. We demonstrate that on synthetic data sets, the proposed method learns more accurate graphs for nonlinearly generated samples; and on benchmark data sets with discrete variables, the learned graphs are reasonably close to the global optima. The code is available at \url{https://github.com/fishmoon1234/DAG-GNN}.

ICML2019. Code is available at https://github.com/fishmoon1234/DAG-GNN

Cited by in corpus (49)

• Iterative Deep Graph Learning for Graph Neural Networks: Better and Robust Node Embeddings
• Causal Discovery with Reinforcement Learning
• D-VAE: A Variational Autoencoder for Directed Acyclic Graphs
• Data-driven learning of robust nonlocal physics from high-fidelity synthetic data
• On the Role of Sparsity and DAG Constraints for Learning Linear DAGs
• A data-driven peridynamic continuum model for upscaling molecular dynamics
• Causality-based CTR Prediction using Graph Neural Networks
• Learning Neural Causal Models from Unknown Interventions
• gCastle: A Python Toolbox for Causal Discovery
• Towards Efficient Local Causal Structure Learning
• Learning Sparse Nonparametric DAGs
• Meta-Weight Graph Neural Network: Push the Limits Beyond Global Homophily
• Discrete Graph Structure Learning for Forecasting Multiple Time Series
• Scalable Causal Structure Learning: Scoping Review of Traditional and Deep Learning Algorithms and New Opportunities in Biomedicine
• Differentiable Causal Discovery Under Unmeasured Confounding
• Causal Discovery from Incomplete Data: A Deep Learning Approach
• Differentiable Causal Discovery from Interventional Data
• Towards Human-like Perception: Learning Structural Causal Model in Heterogeneous Graph
• Learning Neural Causal Models with Active Interventions
• Computably Continuous Reinforcement-Learning Objectives are PAC-learnable
• Gradient-Based Neural DAG Learning
• Graphical Normalizing Flows
• Interactive and Intelligent Root Cause Analysis in Manufacturing with Causal Bayesian Networks and Knowledge Graphs
• Beware of the Simulated DAG! Causal Discovery Benchmarks May Be Easy To Game
• Towards Federated Bayesian Network Structure Learning with Continuous Optimization
• Evaluating structure learning algorithms with a balanced scoring function
• DAG-WGAN: Causal Structure Learning With Wasserstein Generative Adversarial Networks
• Directed Acyclic Graphs With Tears
• FSPN: A New Class of Probabilistic Graphical Model
• Unsuitability of NOTEARS for Causal Graph Discovery
• On the Convergence of Continuous Constrained Optimization for Structure Learning
• A polynomial-time algorithm for learning nonparametric causal graphs
• Generalizable Resource Allocation in Stream Processing via Deep Reinforcement Learning
• Parameter Prediction for Unseen Deep Architectures
• Multi-task Learning of Order-Consistent Causal Graphs
• Methods for Recovering Conditional Independence Graphs: A Survey
• Efficient Neural Causal Discovery without Acyclicity Constraints
• ML4C: Seeing Causality Through Latent Vicinity
• Recursive Bayesian Networks: Generalising and Unifying Probabilistic Context-Free Grammars and Dynamic Bayesian Networks
• Causality Learning With Wasserstein Generative Adversarial Networks
• On the Role of Entropy-based Loss for Learning Causal Structures with Continuous Optimization
• Testing Directed Acyclic Graph via Structural, Supervised and Generative Adversarial Learning
• Gradient-based Causal Structure Learning with Normalizing Flow
• Physical System for Non Time Sequence Data
• Identifiability of AMP chain graph models
• Prequential MDL for Causal Structure Learning with Neural Networks
• VACA: Design of Variational Graph Autoencoders for Interventional and Counterfactual Queries
• DAGSurv: Directed Acyclic Graph Based Survival Analysis Using Deep Neural Networks
• Inductive Granger Causal Modeling for Multivariate Time Series