Deep Structural Causal Models for Tractable Counterfactual Inference
This provides a new approach for answering causal questions in imaging and other domains, addressing a key limitation in existing deep causal learning methods.
The authors tackled the problem of enabling tractable counterfactual inference in structural causal models by integrating deep learning components, achieving successful training on synthetic and real-world medical datasets to handle all levels of Pearl's ladder of causation.
We formulate a general framework for building structural causal models (SCMs) with deep learning components. The proposed approach employs normalising flows and variational inference to enable tractable inference of exogenous noise variables - a crucial step for counterfactual inference that is missing from existing deep causal learning methods. Our framework is validated on a synthetic dataset built on MNIST as well as on a real-world medical dataset of brain MRI scans. Our experimental results indicate that we can successfully train deep SCMs that are capable of all three levels of Pearl's ladder of causation: association, intervention, and counterfactuals, giving rise to a powerful new approach for answering causal questions in imaging applications and beyond. The code for all our experiments is available at https://github.com/biomedia-mira/deepscm.