Supervised Whole DAG Causal Discovery
arXiv:2006.04697
Abstract
We propose to address the task of causal structure learning from data in a supervised manner. Existing work on learning causal directions by supervised learning is restricted to learning pairwise relation, and not well suited for whole DAG discovery. We propose a novel approach of modeling the whole DAG structure discovery as a supervised learning. To fit the problem in hand, we propose to use permutation equivariant models that align well with the problem domain. We evaluate the proposed approach extensively on synthetic graphs of size 10,20,50,100 and real data, and show promising results compared with a variety of previous approaches.
References in corpus (10)
- Adam: A Method for Stochastic Optimization
- Group Equivariant Convolutional Networks
- CAM: Causal additive models, high-dimensional order search and penalized regression
- DAGs with NO TEARS: Continuous Optimization for Structure Learning
- Fashionable Modelling with Flux
- Causal Discovery with Reinforcement Learning
- Deep Models of Interactions Across Sets
- Towards a Learning Theory of Cause-Effect Inference
- Gradient-Based Neural DAG Learning
- The Randomized Causation Coefficient