Learning to target with network interference

2026-05-26

AI Summary

This study addresses the problem of online adaptive target selection under network interference—modeled as spillover effects among individuals—within a sparse linear multi-armed bandit framework. It establishes, for the first time, a unified theoretical framework that accounts for varying levels of prior knowledge about the interference structure: fully known, only the column support size known, or completely unknown. Corresponding lower bounds on regret are derived, and near-optimal structure-aware learning algorithms are developed for each setting. Both theoretical analysis and empirical evaluations demonstrate that the proposed methods significantly outperform baseline approaches that ignore network structure, achieving substantially improved online learning efficiency on both synthetic and real-world datasets.

Abstract

This paper studies adaptive targeting under network interference in a bandit setting, where treatments applied to one individual may affect others through spillover effects. We consider a linear model in a sparse regime, where each individual's outcome can be affected by at most a few others. We first establish a regret lower bound showing that ignoring the network structure and reducing the problem to a standard linear bandit inevitably leads to inefficient learning, particularly in large populations. To understand how structural information can be leveraged, we analyze regimes with varying levels of knowledge of the interference structure: (1) full support knowledge, (2) knowledge of the column support sizes, and (3) no prior knowledge. For each regime, we establish regret lower bounds characterizing the fundamental limits of learning, and develop algorithms that achieve near-optimal regret. Together, our results provide a unified view of how knowledge of the interference structure governs the efficiency of online learning under interference, and offer practical adaptive targeting algorithms in each setting. Numerical experiments on synthetic and real-world data demonstrate the practical benefits of our algorithms.

Problem

• network interference
• adaptive targeting
• spillover effects
• linear bandit
• regret lower bound

Innovation

• network interference
• adaptive targeting
• linear bandits
• spillover effects
• regret bounds