A Graph Autoencoder Approach to Causal Structure Learning

Abstract

Causal structure learning has been a challenging task in the past decades and several mainstream approaches such as constraint- and score-based methods have been studied with theoretical guarantees. Recently, a new approach has transformed the combinatorial structure learning problem into a continuous one and then solved it using gradient-based optimization methods. Following the recent state-of-the-arts, we propose a new gradient-based method to learn causal structures from observational data. The proposed method generalizes the recent gradient-based methods to a graph autoencoder framework that allows nonlinear structural equation models and is easily applicable to vector-valued variables. We demonstrate that on synthetic datasets, our proposed method outperforms other gradient-based methods significantly, especially on large causal graphs. We further investigate the scalability and efficiency of our method, and observe a near linear training time when scaling up the graph size.

NeurIPS 2019 Workshop "Do the right thing": machine learning and causal inference for improved decision making

References in corpus (4)

• Adam: A Method for Stochastic Optimization
• Causal Inference and Causal Explanation with Background Knowledge
• CAM: Causal additive models, high-dimensional order search and penalized regression
• ANAE: Learning Node Context Representation for Attributed Network Embedding

Cited by in corpus (16)

• On the Role of Sparsity and DAG Constraints for Learning Linear DAGs
• Causality-based CTR Prediction using Graph Neural Networks
• gCastle: A Python Toolbox for Causal Discovery
• Scalable Causal Structure Learning: Scoping Review of Traditional and Deep Learning Algorithms and New Opportunities in Biomedicine
• CausalVAE: Structured Causal Disentanglement in Variational Autoencoder
• Learning Neural Causal Models with Active Interventions
• FIRED: a fine-grained robust performance diagnosis framework for cloud applications
• Towards Federated Bayesian Network Structure Learning with Continuous Optimization
• Ordering-Based Causal Discovery with Reinforcement Learning
• Directed Acyclic Graphs With Tears
• On the Convergence of Continuous Constrained Optimization for Structure Learning
• A Bregman Method for Structure Learning on Sparse Directed Acyclic Graphs
• On Low Rank Directed Acyclic Graphs and Causal Structure Learning
• Causal Inference in medicine and in health policy, a summary
• On the Role of Entropy-based Loss for Learning Causal Structures with Continuous Optimization
• Inference of Causal Effects when Control Variables are Unknown