Document Type
Discussion Paper
Publication Date
12-1-2018
CFDP Number
2155
CFDP Pages
15
Journal of Economic Literature (JEL) Code(s)
C1, C5, C9
Abstract
What is the most statistically efficient way to do off-policy optimization with batch data from bandit feedback? For log data generated by contextual bandit algorithms, we consider offline estimators for the expected reward from a counterfactual policy. Our estimators are shown to have lowest variance in a wide class of estimators, achieving variance reduction relative to standard estimators. We then apply our estimators to improve advertisement design by a major advertisement company. Consistent with the theoretical result, our estimators allow us to improve on the existing bandit algorithm with more statistical confidence compared to a state-of-theart benchmark.
Recommended Citation
Narita, Yusuke; Yasui, Shota; and Yata, Kohei, "Efficient Counterfactual Learning from Bandit Feedback" (2018). Cowles Foundation Discussion Papers. 110.
https://elischolar.library.yale.edu/cowles-discussion-paper-series/110