From Statistical to Causal Learning
It addresses foundational problems in AI for researchers, but is incremental as it reviews existing ideas without presenting new results.
The paper discusses the evolution of AI research from symbolic and statistical methods to causal models, highlighting that key challenges in machine learning and AI are tied to causality, and progress depends on better modeling and inference techniques.
We describe basic ideas underlying research to build and understand artificially intelligent systems: from symbolic approaches via statistical learning to interventional models relying on concepts of causality. Some of the hard open problems of machine learning and AI are intrinsically related to causality, and progress may require advances in our understanding of how to model and infer causality from data.