Counterfactual Data-Fusion for Online Reinforcement Learners
Andrew Forney, Judea Pearl, Elias Bareinboim
Abstract
The Multi-Armed Bandit problem with Unobserved Confounders (MABUC) considers decision-making settings where unmeasured variables can influence both the agent’s decisions and received rewards (Bareinboim et al., 2015). Recent findings showed that unobserved confounders (UCs) pose a unique challenge to algorithms based on standard randomization (i.e., experimental data); if UCs are naively averaged out, these algorithms behave sub-optimally, possibly incurring infinite regret. In this paper, we show how counterfactual-based decision-making circumvents these problems and leads to a coherent fusion of observational and experimental data. We then demonstrate this new strategy in an enhanced Thompson Sampling bandit player, and support our findings’ efficacy with extensive simulations.
BibTeX
@InProceedings{pmlr-v70-forney17a,
title = {Counterfactual Data-Fusion for Online Reinforcement Learners},
author = {Andrew Forney and Judea Pearl and Elias Bareinboim},
booktitle = {Proceedings of the 34th International Conference on Machine Learning},
pages = {1156--1164},
year = {2017},
editor = {Precup, Doina and Teh, Yee Whye},
volume = {70},
series = {Proceedings of Machine Learning Research},
month = {06--11 Aug},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v70/forney17a/forney17a.pdf},
url = {https://proceedings.mlr.press/v70/forney17a.html},
abstract = {The Multi-Armed Bandit problem with Unobserved Confounders (MABUC) considers decision-making settings where unmeasured variables can influence both the agent’s decisions and received rewards (Bareinboim et al., 2015). Recent findings showed that unobserved confounders (UCs) pose a unique challenge to algorithms based on standard randomization (i.e., experimental data); if UCs are naively averaged out, these algorithms behave sub-optimally, possibly incurring infinite regret. In this paper, we show how counterfactual-based decision-making circumvents these problems and leads to a coherent fusion of observational and experimental data. We then demonstrate this new strategy in an enhanced Thompson Sampling bandit player, and support our findings’ efficacy with extensive simulations.}
}