A comparative study of counterfactual estimators
By Thomas Nedelec, Nicolas Le Roux, Vianney Perchet. 2017
10 citations recorded by the source corpus.
We provide a comparative study of several widely used off-policy estimators (Empirical Average, Basic Importance Sampling and Normalized Importance Sampling), detailing the different regimes where they are individually suboptimal. We then exhibit properties optimal estimators should possess. In the case where examples have been gathered using multiple policies, we show that fused estimators dominate basic ones but can still be improved.
Subfields: Economics and Econometrics, Statistics and Probability
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Economics and Econometrics
- Jean-Baptiste Lespiau
- Theophane Weber
- Fernando Diaz
- Mounia Lalmas
- Nicolas Manfred Otto Heess
- Arthur Guez
In Statistics and Probability
- Theophane Weber
- Fernando Diaz
- Mounia Lalmas
- Lars Buesing
- Sébastien Racanière
- James McInerney
- Nicolas Manfred Otto Heess
- Arthur Guez
Citations in the corpus
- Woulda, Coulda, Shoulda: Counterfactually-Guided Policy Search (2018, background): “However, their and related approaches (Li et al., 2015; Jiang & Li, 2015; Swaminathan & Joachims, 2015; Nedelec et al., 2017; Atan et al., 2016; Thomas & Brunskill, 2016) such as doubly-robust estimators, rely on Importance Sampling (IS), also called Propensity Score method.”
- Towards a Fair Marketplace: Counterfactual Evaluation of the trade-off between Relevance, Fairness & Satisfaction in Recommendation Systems (2018, background): “Nedelec et al. [20] present a detailed overview and comparison of many different counterfactual estimators.”