Causal Deepsets for Off-policy Evaluation under Spatial or Spatio-temporal Interferences

TL;DR

This paper proposes a causal deepset framework for off-policy evaluation that relaxes traditional assumptions, allowing for more flexible and accurate estimation in complex spatio-temporal interference settings.

Contribution

It introduces a permutation invariance assumption and algorithms that adaptively learn the mean-field function, enhancing OPE accuracy beyond existing methods.

Findings

More precise estimations than baseline algorithms.

Theoretical validation of the new algorithms.

Improved practical applicability of OPE methods.

Abstract

Off-policy evaluation (OPE) is widely applied in sectors such as pharmaceuticals and e-commerce to evaluate the efficacy of novel products or policies from offline datasets. This paper introduces a causal deepset framework that relaxes several key structural assumptions, primarily the mean-field assumption, prevalent in existing OPE methodologies that handle spatio-temporal interference. These traditional assumptions frequently prove inadequate in real-world settings, thereby restricting the capability of current OPE methods to effectively address complex interference effects. In response, we advocate for the implementation of the permutation invariance (PI) assumption. This innovative approach enables the data-driven, adaptive learning of the mean-field function, offering a more flexible estimation method beyond conventional averaging. Furthermore, we present novel algorithms that incorporate the PI assumption into OPE and thoroughly examine their theoretical foundations. Our numerical analyses demonstrate that this novel approach yields significantly more precise estimations than existing baseline algorithms, thereby substantially improving the practical applicability and effectiveness of OPE methodologies. A Python implementation of our proposed method is available at https://github.com/BIG-S2/Causal-Deepsets.