Toward Causal Representation Learning

The two fields of machine learning and graphical causality arose and are developed separately. However, there is, now, cross-pollination and increasing interest in both fields to benefit from the advances of the other. In this article, we review fundamental concepts of causal inference and relate them to crucial open problems of machine learning, including transfer and generalization, thereby assaying how causality can contribute to modern machine learning research. This also applies in the opposite direction: we note that most work in causality starts from the premise that the causal variables are given. A central problem for AI and causality is, thus, causal representation learning, that is, the discovery of high-level causal variables from low-level observations. Finally, we delineate some implications of causality for machine learning and propose key research areas at the intersection of both communities.

Automated summary

The fields of machine learning and graphical causality have started to influence each other and show interest in benefiting from each other's advancements. Understanding fundamental concepts of causal inference, and relating them to key issues in machine learning, can help enhance modern machine learning research. A central problem in the intersection of AI and causality is the learning of causal representations, which involves discovering high-level causal variables from low-level observations.