CAusal modeling is an umbrella time period for a variety of strategies that permit us to mannequin the results of our actions on the world.
Causal fashions differ from conventional machine studying fashions in some ways.
Crucial distinction between them is that the knowledge contained within the observational information used to coach conventional machine studying machines is mostly inadequate to persistently mannequin the results of our actions. It comes from the truth that
outcome?
Utilizing conventional machine studying strategies to mannequin the outcomes of actions nearly at all times results in biased choices.
instance right here is utilizing a regression mannequin educated on historic information to find out advertising combine.
One other one?
Use XGBoost educated on previous observations to foretell the chance of churn and ship a marketing campaign if the anticipated chance of churn is larger than some threshold.
