Exploring and exploiting the seemingly harmless theorem behind Double Machine Studying

Causal inference, and particularly causal machine studying, is an indispensable software that may assist us make selections by understanding trigger and impact. Optimizing costs, lowering buyer churn, operating focused advert campaigns, and deciding which sufferers would profit most from medical remedy are all instance use circumstances for causal machine studying.

There are a lot of methods for causal machine studying issues, however the method that appears to face out most is called Double Machine Studying (DML) or Debiased/Orthogonal Machine Studying. Past the empirical success of DML, this method stands out due to its wealthy theoretical backing rooted in a easy theorem from econometrics.

On this article, we’ll unpack the theory that grounds DML via hands-on examples. We’ll focus on the instinct for DML and empirically confirm its generality on more and more advanced examples. This text just isn’t a tutorial on DML, as a substitute it serves as motivation for the way DML fashions see previous mere correlation to know and predict trigger and impact.