Gradient-Based Neural DAG Learning

Abstract

We propose a novel score-based approach to learning a directed acyclic graph (DAG) from observational data. We adapt a recently proposed continuous constrained optimization formulation to allow for nonlinear relationships between variables using neural networks. This extension allows to model complex interactions while avoiding the combinatorial nature of the problem. In addition to comparing our method to existing continuous optimization methods, we provide missing empirical comparisons to nonlinear greedy search methods. On both synthetic and real-world data sets, this new method outperforms current continuous methods on most tasks, while being competitive with existing greedy search methods on important metrics for causal inference.

arXiv commentAppears in: Proceedings of the Eighth International Conference on Learning Representations (ICLR 2020). 23 pages

References in corpus (16)

• Emergence of scaling in random networks
• MADE: Masked Autoencoder for Distribution Estimation
• On the Identifiability of the Post-Nonlinear Causal Model
• CAM: Causal additive models, high-dimensional order search and penalized regression
• Identifiability of Gaussian structural equation models with equal error variances
• DAGs with NO TEARS: Continuous Optimization for Structure Learning
• Bayesian network learning with cutting planes
• Causal Inference on Discrete Data using Additive Noise Models
• DAG-GNN: DAG Structure Learning with Graph Neural Networks
• ${\ell }_0$-penalized maximum likelihood for sparse directed acyclic graphs
• High-dimensional learning of linear causal networks via inverse covariance estimation
• Domain Adaptation by Using Causal Inference to Predict Invariant Conditional Distributions
• Theory of Deep Learning III: explaining the non-overfitting puzzle
• Learning Directed Acyclic Graphs with Penalized Neighbourhood Regression
• Structural Agnostic Modeling: Adversarial Learning of Causal Graphs
• Structural Intervention Distance (SID) for Evaluating Causal Graphs

Cited by in corpus (12)

• Causal Discovery with Reinforcement Learning
• Causal Induction from Visual Observations for Goal Directed Tasks
• Learning Neural Causal Models from Unknown Interventions
• Learning Sparse Nonparametric DAGs
• Graphical Normalizing Flows
• Counterfactual Data Augmentation using Locally Factored Dynamics
• Causal Adversarial Network for Learning Conditional and Interventional Distributions
• A polynomial-time algorithm for learning nonparametric causal graphs
• Causal Discovery from Incomplete Data using An Encoder and Reinforcement Learning
• Supervised Whole DAG Causal Discovery
• Efficient and Scalable Structure Learning for Bayesian Networks: Algorithms and Applications
• Relate and Predict: Structure-Aware Prediction with Jointly Optimized Neural DAG