Learning and Decision-Making with Data: Optimal Formulations and Phase Transitions

TL;DR

This paper investigates optimal data-driven decision-making formulations based on historical data, revealing phase transitions between different regimes and connecting robust, entropic, and variance penalized approaches.

Contribution

It introduces a novel framework for designing optimal data-driven formulations by measuring their out-of-sample performance and characterizes phase transitions among three regimes.

Findings

Identifies three distinct out-of-sample performance regimes.

Shows the existence of phase transitions between regimes.

Connects robust, entropic, and variance penalized formulations.

Abstract

We study the problem of designing optimal learning and decision-making formulations when only historical data is available. Prior work typically commits to a particular class of data-driven formulation and subsequently tries to establish out-of-sample performance guarantees. We take here the opposite approach. We define first a sensible yard stick with which to measure the quality of any data-driven formulation and subsequently seek to find an optimal such formulation. Informally, any data-driven formulation can be seen to balance a measure of proximity of the estimated cost to the actual cost while guaranteeing a level of out-of-sample performance. Given an acceptable level of out-of-sample performance, we construct explicitly a data-driven formulation that is uniformly closer to the true cost than any other formulation enjoying the same out-of-sample performance. We show the existence of three distinct out-of-sample performance regimes (a superexponential regime, an exponential regime and a subexponential regime) between which the nature of the optimal data-driven formulation experiences a phase transition. The optimal data-driven formulations can be interpreted as a classically robust formulation in the superexponential regime, an entropic distributionally robust formulation in the exponential regime and finally a variance penalized formulation in the subexponential regime. This final observation unveils a surprising connection between these three, at first glance seemingly unrelated, data-driven formulations which until now remained hidden.