Clusters
Propensity Score
The propensity score is observable:
\[\pi(X_i) = \mathbb{P}(D_i = 1 \mid X_i)\]
That is, given access to the following probability space, we can compute the probability measure.
\[\big( \mathcal{X}\times \mathcal{D} \times \mathcal{Y}, \mathcal{F}, \mathcal{P} \big)\]
To Do
Add in the details