Positivity-free Policy Learning with Observational Data

TL;DR

This paper introduces a positivity-free policy learning framework that effectively learns optimal treatment policies from observational data without relying on the positivity assumption, using incremental propensity scores and semiparametric efficiency theory.

Contribution

The study proposes a novel positivity-free policy learning method with incremental propensity scores, providing theoretical guarantees and efficient estimators applicable with machine learning techniques.

Findings

Achieves rapid convergence rates with efficient estimators.

Validates the framework's finite-sample performance through numerical experiments.

Ensures robust causal effect identification without positivity assumption.

Abstract

Policy learning utilizing observational data is pivotal across various domains, with the objective of learning the optimal treatment assignment policy while adhering to specific constraints such as fairness, budget, and simplicity. This study introduces a novel positivity-free (stochastic) policy learning framework designed to address the challenges posed by the impracticality of the positivity assumption in real-world scenarios. This framework leverages incremental propensity score policies to adjust propensity score values instead of assigning fixed values to treatments. We characterize these incremental propensity score policies and establish identification conditions, employing semiparametric efficiency theory to propose efficient estimators capable of achieving rapid convergence rates, even when integrated with advanced machine learning algorithms. This paper provides a thorough exploration of the theoretical guarantees associated with policy learning and validates the proposed framework's finite-sample performance through comprehensive numerical experiments, ensuring the identification of causal effects from observational data is both robust and reliable.