Adaptive Experimental Design Using Shrinkage Estimators

TL;DR

This paper introduces an adaptive experimental design for multi-armed trials that employs a Stein-like shrinkage estimator to improve causal effect estimation by borrowing information across treatments, leading to reduced estimation error.

Contribution

It proposes a novel adaptive treatment assignment method using a Stein-like shrinkage estimator for heteroscedastic data, enhancing efficiency over traditional Neyman allocation schemes.

Findings

Shrinkage estimator reduces expected squared error compared to isolated estimates.

The expected loss can be computed efficiently via Gaussian quadratic form.

Simulations show the method significantly improves estimation accuracy.

Abstract

In the setting of multi-armed trials, adaptive designs are a popular way to increase estimation efficiency, identify optimal treatments, or maximize rewards to individuals. Recent work has considered the case of estimating the effects of K active treatments, relative to a control arm, in a sequential trial. Several papers have proposed sequential versions of the classical Neyman allocation scheme to assign treatments as individuals arrive, typically with the goal of using Horvitz-Thompson-style estimators to obtain causal estimates at the end of the trial. However, this approach may be inefficient in that it fails to borrow information across the treatment arms. In this paper, we consider adaptivity when the final causal estimation is obtained using a Stein-like shrinkage estimator for heteroscedastic data. Such an estimator shares information across treatment effect estimates, providing provable reductions in expected squared error loss relative to estimating each causal effect in isolation. Moreover, we show that the expected loss of the shrinkage estimator takes the form of a Gaussian quadratic form, allowing it to be computed efficiently using numerical integration. This result paves the way for sequential adaptivity, allowing treatments to be assigned to minimize the shrinker loss. Through simulations, we demonstrate that this approach can yield meaningful reductions in estimation error. We also characterize how our adaptive algorithm assigns treatments differently than would a sequential Neyman allocation.