Resolving Spurious Correlations in Causal Models of Environments via Interventions

Causality (Halpern & Pearl, 2005) is an important concept (Pearl, 2018) for Machine Learning, since it resolves many issues in performance and Artificial Intelligence (AI) safety (Amodei et al., 2016) such as interpretability (Madumal et al., 2019; Bengio, 2017), robustness to distributional shift (de Haan et al., 2019a) and sample-efficiency (Buesing et al., 2018). It is particularly well suited for Reinforcement Learning (RL), compared to supervised learning, because in RL there is an opportunity to take actions and influence the environment in a directed way. Since causality is a cornerstone in science, such an agent is expected to be superior to noncausal agents (Marino et al., 2019). Spurious correlations are a major obstacle in learning causal models. If present, they make learning from purely observational data impossible (Pearl & Mackenzie, 2018). We take advantage of the fact that it is possible to uncover the causal graph by executing interventions (Halpern & Pearl, 2005) which change the data distribution. We design a method to automatically resolve spurious correlations when learning the causal graph of the environment.