Balanced Off-Policy Evaluation for Personalized Pricing

TL;DR

This paper introduces a balanced off-policy evaluation method tailored for personalized pricing, addressing challenges of deterministic logging policies by minimizing worst-case mean squared error.

Contribution

It proposes a new approach based on balanced policy evaluation that improves off-policy evaluation accuracy in pricing scenarios with limited exploration.

Findings

The method outperforms inverse propensity weighting in deterministic settings.

Theoretical guarantees ensure convergence of the proposed estimator.

Empirical results on real-world data demonstrate improved evaluation accuracy.

Abstract

We consider a personalized pricing problem in which we have data consisting of feature information, historical pricing decisions, and binary realized demand. The goal is to perform off-policy evaluation for a new personalized pricing policy that maps features to prices. Methods based on inverse propensity weighting (including doubly robust methods) for off-policy evaluation may perform poorly when the logging policy has little exploration or is deterministic, which is common in pricing applications. Building on the balanced policy evaluation framework of Kallus (2018), we propose a new approach tailored to pricing applications. The key idea is to compute an estimate that minimizes the worst-case mean squared error or maximizes a worst-case lower bound on policy performance, where in both cases the worst-case is taken with respect to a set of possible revenue functions. We establish theoretical convergence guarantees and empirically demonstrate the advantage of our approach using a real-world pricing dataset.