Counterfactual Learning of Stochastic Policies with Continuous Actions

TL;DR

This paper advances counterfactual learning for continuous actions by introducing a kernel embedding model, emphasizing optimization techniques, and proposing an evaluation protocol with a new dataset for real-world applications.

Contribution

It presents a novel kernel embedding approach for continuous actions, explores optimization methods, and develops an evaluation protocol with a large-scale dataset.

Findings

Kernel embedding model improves counterfactual estimation

Proximal algorithms enhance optimization stability

New dataset enables realistic offline policy evaluation

Abstract

Counterfactual reasoning from logged data has become increasingly important for many applications such as web advertising or healthcare. In this paper, we address the problem of learning stochastic policies with continuous actions from the viewpoint of counterfactual risk minimization (CRM). While the CRM framework is appealing and well studied for discrete actions, the continuous action case raises new challenges about modelization, optimization, and~offline model selection with real data which turns out to be particularly challenging. Our paper contributes to these three aspects of the CRM estimation pipeline. First, we introduce a modelling strategy based on a joint kernel embedding of contexts and actions, which overcomes the shortcomings of previous discretization approaches. Second, we empirically show that the optimization aspect of counterfactual learning is important, and we demonstrate the benefits of proximal point algorithms and smooth estimators. Finally, we propose an evaluation protocol for offline policies in real-world logged systems, which is challenging since policies cannot be replayed on test data, and we release a new large-scale dataset along with multiple synthetic, yet realistic, evaluation setups.