In this talk, I will talk about causal inference for time series data. Our focus will be on discrete-time observations, and considering various types of causal graphs, from detailed ones where each node represents a time series at specific time points, to more abstract representations where each node is an entire time series.
I will discuss causal discovery, the process of inferring causal graphs from observational data, as well as causal reasoning, specifically the challenge of identifying the total effect of interventions when only abstracted versions of the true causal graph are available.
