Synthesis with casual inference and reinforcement learning

[isaacmg]

Since ultimately the goal of time series forecasting is to provide actionable insights to solve a problem I think in the long term it makes sense to incorporate casual inference and RL based methods. Though this would likely require additional data sources of various interventions/policies etc...