Learn the way to utilize DML for estimating individual level treatment effects to enable data-driven targeting

This text is the 2nd in a 2 part series on simplifying and democratizing Double Machine Learning. Within the 1st part, we covered the basics of Double Machine Learning, together with two basic causal inference applications. Now, in pt. 2, we’ll extend this data to show our causal inference problem right into a prediction task, wherein we predict individual level treatment effects to help in decision making and data-driven targeting.

Double Machine Learning, as we learned in part 1 of this series, is a highly flexible partially-linear causal inference method for estimating the typical treatment effect (ATE) of a treatment. Specifically, it might probably be utilized to model highly non-linear confounding relationships in observational data and/or to cut back the variation in our key final result in experimental settings. Estimating the ATE is especially useful in understanding the typical impact of a selected treatment, which might be extremely useful for future decision making. Nonetheless, extrapolating this treatment effect assumes a level homogeneity within the effect; that’s, whatever the population we roll treatment out to, we anticipate the effect to be just like the ATE. What if we’re limited within the number of people who we will goal for future rollout and thus want to grasp amongst which subpopulations the treatment was only to drive highly effective rollout?

This issue described above concerns estimating treatment effect heterogeneity. That’s, how does our treatment effect impact different subsets of the population? Luckily for us, DML provides a strong framework to do exactly this. Specifically, we will make use of DML to estimate the Conditional Average Treatment Effect (CATE). First, let’s revisit our definition of the ATE:

Now with the CATE, we estimate the ATE conditional on a set of values for our covariates, X: