Learning Optimal Policies from Observational Data
By Onur Atan, William R. Zame, Mihaela van der Schaar. 2018
20 citations recorded by the source corpus.
Choosing optimal (or at least better) policies is an important problem in domains from medicine to education to finance and many others. One approach to this problem is through controlled experiments/trials - but controlled experiments are expensive. Hence it is important to choose the best policies on the basis of observational data. This presents two difficult challenges: (i) missing counterfactuals, and (ii) selection bias. This paper presents theoretical bounds on estimation errors of counterfactuals from observational data by making connections to domain adaptation theory. It also presents a principled way of choosing optimal policies using domain adversarial neural networks. We illustrate the effectiveness of domain adversarial training together with various features of our algorithm on a semi-synthetic breast cancer dataset and a supervised UCI dataset (Statlog).
Subfields: Artificial Intelligence, 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 Artificial Intelligence
- Mihaela van der Schaar
- Ahmed M. Alaa
- Ioana Bica
- James Jordon
- Yaliang Li
- Jing Gao
- Aidong Zhang
- Sheng Li
In Statistics and Probability
- Mihaela van der Schaar
- Ahmed M. Alaa
- Ioana Bica
- James Jordon
- Yaliang Li
- Jing Gao
- Aidong Zhang
- Sheng Li
Citations in the corpus
- Estimating Counterfactual Treatment Outcomes over Time Through Adversarially Balanced Representations (2020, methodology): “Atan et al. (2018) only considers the static setting and aims to choose the optimal treatment instead of estimating all of the counterfactual outcomes.”
- A Survey on Causal Inference (2020, methodology): “Later, various methods [10, 19, 74, 88, 154, 155, 158, 179] are proposed to improve those two methods with different settings.”
- Constructing effective personalized policies using counterfactual inference from biased data sets with many features (2016, background): “Our own (preliminary) work (Atan et al. 2018) propose a different approach for learning a representation function and a policy.”