A step towards the applicability of algorithms based on invariant causal learning on observational data
This work addresses a bottleneck for researchers using causal learning in real-world applications where multiple environments are lacking, but it is incremental as it builds on existing methods like ICP.
The paper tackles the problem of applying invariant causal learning algorithms to observational data by proposing a method to generate multiple training environments, which are normally unavailable, and demonstrates its performance with Invariant Causal Prediction on simulated data.
Machine learning can benefit from causal discovery for interpretation and from causal inference for generalization. In this line of research, a few invariant learning algorithms for out-of-distribution (OOD) generalization have been proposed by using multiple training environments to find invariant relationships. Some of them are focused on causal discovery as Invariant Causal Prediction (ICP), which finds causal parents of a variable of interest, and some directly provide a causal optimal predictor that generalizes well in OOD environments as Invariant Risk Minimization (IRM). This group of algorithms works under the assumption of multiple environments that represent different interventions in the causal inference context. Those environments are not normally available when working with observational data and real-world applications. Here we propose a method to generate them in an efficient way. We assess the performance of this unsupervised learning problem by implementing ICP on simulated data. We also show how to apply ICP efficiently integrated with our method for causal discovery. Finally, we proposed an improved version of our method in combination with ICP for datasets with multiple covariates where ICP and other causal discovery methods normally degrade in performance.