Learning Causal Models under Independent Changes

In many scientific applications, we observe a system in different conditions in which its components may change, rather than in isolation.