Off-Policy Evaluation and Learning for External Validity under a Covariate Shift

Masatoshi Uehara, Masahiro Kato, and Shota Yasui (equal contribution)

Advances in Neural Information Processing Systems (NeurIPS), 2020. Spotlight (top 3%, 280/9054).

In one sentence. Off-policy evaluation and learning when the population you will deploy on differs from the population that generated the logs: the paper derives the efficiency bound under covariate shift and gives doubly robust, efficient estimators.

The external validity gap

Off-policy evaluation estimates the expected reward of a new policy from data logged under a different policy, and off-policy learning searches for the policy that maximizes it. Both standardly assume that the covariate distribution of the historical data matches that of the population where the new policy will run.

That assumption is frequently false — the logs come from last quarter's users, or from one market, and the policy is meant for another. This is a covariate shift, and it is exactly the external validity problem in econometrics stated in a bandit vocabulary.

Contributions

Related

See the full publication list.