Universal Off-Policy Evaluation

04/26/2021
by   Yash Chandak, et al.
0

When faced with sequential decision-making problems, it is often useful to be able to predict what would happen if decisions were made using a new policy. Those predictions must often be based on data collected under some previously used decision-making rule. Many previous methods enable such off-policy (or counterfactual) estimation of the expected value of a performance measure called the return. In this paper, we take the first steps towards a universal off-policy estimator (UnO) – one that provides off-policy estimates and high-confidence bounds for any parameter of the return distribution. We use UnO for estimating and simultaneously bounding the mean, variance, quantiles/median, inter-quantile range, CVaR, and the entire cumulative distribution of returns. Finally, we also discuss Uno's applicability in various settings, including fully observable, partially observable (i.e., with unobserved confounders), Markovian, non-Markovian, stationary, smoothly non-stationary, and discrete distribution shifts.

READ FULL TEXT

page 1

page 2

page 3

page 4

01/25/2021

High-Confidence Off-Policy (or Counterfactual) Variance Estimation

Many sequential decision-making systems leverage data collected using pr...
02/25/2022

Decision Making in Non-Stationary Environments with Policy-Augmented Monte Carlo Tree Search

Decision-making under uncertainty (DMU) is present in many important pro...
11/28/2021

Identification of Subgroups With Similar Benefits in Off-Policy Policy Evaluation

Off-policy policy evaluation methods for sequential decision making can ...
10/23/2020

Towards Safe Policy Improvement for Non-Stationary MDPs

Many real-world sequential decision-making problems involve critical sys...
04/18/2021

Off-Policy Risk Assessment in Contextual Bandits

To evaluate prospective contextual bandit policies when experimentation ...
03/12/2020

Off-policy Policy Evaluation For Sequential Decisions Under Unobserved Confounding

When observed decisions depend only on observed features, off-policy pol...
12/16/2020

Lévy walks derived from a Bayesian decision-making model in non-stationary environments

Lévy walks are found in the migratory behaviour patterns of various orga...