What is the Right Notion of Distance between Predict-then-Optimize Tasks?

TL;DR

This paper introduces OTD$^3$, a new dataset distance measure tailored for Predict-then-Optimize tasks that considers downstream decisions, improving the assessment of dataset similarity and transferability.

Contribution

The paper proposes OTD$^3$, a novel decision-aware dataset distance for PtO tasks, and demonstrates its effectiveness over traditional distances in predicting model transferability.

Findings

OTD$^3$ better predicts transferability across PtO tasks.

Traditional feature-label distances are less informative in PtO settings.

Derived a PtO-specific adaptation bound based on OTD$^3$.

Abstract

Comparing datasets is a fundamental task in machine learning, essential for various learning paradigms-from evaluating train and test datasets for model generalization to using dataset similarity for detecting data drift. While traditional notions of dataset distances offer principled measures of similarity, their utility has largely been assessed through prediction error minimization. However, in Predict-then-Optimize (PtO) frameworks, where predictions serve as inputs for downstream optimization tasks, model performance is measured through decision regret rather than prediction error. In this work, we propose OTD$^3$ (Optimal Transport Decision-aware Dataset Distance), a novel dataset distance that incorporates downstream decisions in addition to features and labels. We show that traditional feature-label distances lack informativeness in PtO settings, while OTD$^3$ more effectively captures adaptation success. We also derive a PtO-specific adaptation bound based on this distance. Empirically, we show that our proposed distance accurately predicts model transferability across three different PtO tasks from the literature. The code is available at https://github.com/paularodr/OTD3.