Optimal and Adaptive Off-policy Evaluation in Contextual Bandits
Published in
International Conference on Machine Learning(2016)External Links:
TL
TL;DR
The SWITCH estimator is proposed, which can use an existing reward model to achieve a better bias-variance tradeoff than IPS and DR and prove an upper bound on its MSE and demonstrate its benefits empirically on a diverse collection of data sets, often outperforming prior work by orders of magnitude.
Abstract
We study the off-policy evaluation problem---estimating the value of a target policy using data collected by another policy---under the contextual bandit model. We consider the general (agnostic) setting without access to a consistent model of rewards and establish a minimax lower bound on the mean squared error (MSE). The bound is matched up to constants by the inverse propensity scoring (IPS) and doubly robust (DR) estimators. This highlights the difficulty of the agnostic contextual setting, in contrast with multi-armed bandits and contextual bandits with access to a consistent reward model, where IPS is suboptimal. We then propose the SWITCH estimator, which can use an existing reward model (not necessarily consistent) to achieve a better bias-variance tradeoff than IPS and DR. We prove an upper bound on its MSE and demonstrate its benefits empirically on a diverse collection of data sets, often outperforming prior work by orders of magnitude.
References32 items
1
The Value of Knowing the Propensity Score for Estimating Average Treatment Effects
2
Data-Efficient Off-Policy Policy Evaluation for Reinforcement Learning
3
Doubly Robust Off-policy Value Evaluation for Reinforcement Learning
4
Toward Minimax Off-policy Value Estimation
5