Efficient Neural Causal Discovery without Acyclicity Constraints

Abstract

Learning the structure of a causal graphical model using both observational and interventional data is a fundamental problem in many scientific fields. A promising direction is continuous optimization for score-based methods, which, however, require constrained optimization to enforce acyclicity or lack convergence guarantees. In this paper, we present ENCO, an efficient structure learning method for directed, acyclic causal graphs leveraging observational and interventional data. ENCO formulates the graph search as an optimization of independent edge likelihoods, with the edge orientation being modeled as a separate parameter. Consequently, we can provide convergence guarantees of ENCO under mild conditions without constraining the score function with respect to acyclicity. In experiments, we show that ENCO can efficiently recover graphs with hundreds of nodes, an order of magnitude larger than what was previously possible, while handling deterministic variables and latent confounders.

arXiv commentPublished as a conference paper at the International Conference on Learning Representations (ICLR), 2022

References in corpus (19)

• Characterization and Greedy Learning of Interventional Markov Equivalence Classes of Directed Acyclic Graphs
• A Survey of Learning Causality with Data: Problems and Methods
• DAGs with NO TEARS: Continuous Optimization for Structure Learning
• DAG-GNN: DAG Structure Learning with Graph Neural Networks
• On Causal and Anticausal Learning
• A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms
• Causal Discovery with Reinforcement Learning
• Variational Autoencoder with Arbitrary Conditioning
• Permutation-based Causal Inference Algorithms with Interventions
• On the Role of Sparsity and DAG Constraints for Learning Linear DAGs
• Learning Neural Causal Models from Unknown Interventions
• Characterizing and Learning Equivalence Classes of Causal DAGs under Interventions
• Almost Optimal Intervention Sets for Causal Discovery
• On the Number of Experiments Sufficient and in the Worst Case Necessary to Identify All Causal Relations Among N Variables
• Learning Sparse Nonparametric DAGs
• Causal Generative Neural Networks
• Differentiable Causal Discovery from Interventional Data
• Causal Discovery with General Non-Linear Relationships Using Non-Linear ICA
• Learning DAGs without imposing acyclicity

Cited by in corpus (1)

• Learning Neural Causal Models with Active Interventions