Causal inference, and specifically causal machine learning, is an indispensable tool that can help us make decisions by understanding cause and effect. Optimizing prices, reducing customer churn, running targeted ad campaigns, and deciding which patients would benefit most from medical treatment are all example use cases for causal machine learning.
There are many techniques for causal machine learning problems, but the technique that seems to stand out most is known as Double Machine Learning (DML) or Debiased/Orthogonal Machine Learning. Beyond the empirical success of DML, this technique stands out because of its rich theoretical backing rooted in a simple theorem from econometrics.
In this article, we’ll unpack the theorem that grounds DML through hands-on examples. We’ll discuss the intuition for DML and empirically verify its generality on increasingly complex examples. This article is not a tutorial on DML, instead it serves as motivation for how DML models see past mere correlation to understand and predict cause and effect.