Autoregressive flow-based causal discovery and inference
This work addresses causal inference challenges for researchers and practitioners in machine learning and statistics, offering a flexible approach that reduces restrictive assumptions, though it appears incremental by building on existing flow models.
The authors tackled the problem of performing causal inference tasks, including causal discovery and making interventional and counterfactual predictions, by proposing an autoregressive flow-based method that leverages the ordering of variables and invertibility of flows, achieving favorable results on synthetic data and the Cause-Effect Pairs benchmark dataset.
We posit that autoregressive flow models are well-suited to performing a range of causal inference tasks - ranging from causal discovery to making interventional and counterfactual predictions. In particular, we exploit the fact that autoregressive architectures define an ordering over variables, analogous to a causal ordering, in order to propose a single flow architecture to perform all three aforementioned tasks. We first leverage the fact that flow models estimate normalized log-densities of data to derive a bivariate measure of causal direction based on likelihood ratios. Whilst traditional measures of causal direction often require restrictive assumptions on the nature of causal relationships (e.g., linearity),the flexibility of flow models allows for arbitrary causal dependencies. Our approach compares favourably against alternative methods on synthetic data as well as on the Cause-Effect Pairs bench-mark dataset. Subsequently, we demonstrate that the invertible nature of flows naturally allows for direct evaluation of both interventional and counterfactual predictions, which require marginalization and conditioning over latent variables respectively. We present examples over synthetic data where autoregressive flows, when trained under the correct causal ordering, are able to make accurate interventional and counterfactual predictions