Off-Policy Evaluation via Adaptive Weighting with Data from Contextual Bandits
This work addresses the challenge of evaluating new policies from adaptively collected data, which is crucial for applications like online advertising and healthcare, though it is incremental as it builds on existing doubly robust estimators.
The paper tackles the problem of off-policy evaluation from contextual bandit data, where existing estimators like doubly robust ones suffer from high variance due to exploding importance weights. It proposes an adaptive weighting method that improves accuracy and enables valid confidence intervals, as demonstrated with synthetic data and benchmarks.
It has become increasingly common for data to be collected adaptively, for example using contextual bandits. Historical data of this type can be used to evaluate other treatment assignment policies to guide future innovation or experiments. However, policy evaluation is challenging if the target policy differs from the one used to collect data, and popular estimators, including doubly robust (DR) estimators, can be plagued by bias, excessive variance, or both. In particular, when the pattern of treatment assignment in the collected data looks little like the pattern generated by the policy to be evaluated, the importance weights used in DR estimators explode, leading to excessive variance. In this paper, we improve the DR estimator by adaptively weighting observations to control its variance. We show that a t-statistic based on our improved estimator is asymptotically normal under certain conditions, allowing us to form confidence intervals and test hypotheses. Using synthetic data and public benchmarks, we provide empirical evidence for our estimator's improved accuracy and inferential properties relative to existing alternatives.