Strategic Decision Focused Learning
Strategic decision-focused learning predicts exogenous states for multi-agent games where better accuracy can reduce equilibrium payoffs, requiring strategic-aware predictors.
Published 2026Atlanta Poster Session 2 · Wed, Dec 9, 4:30 PM–7:30 PM local time · Hall C1arXiv ↗OpenReview ↗
Only vote on papers you've read. Sign in with GitHub to vote.
A rigorous, must-read formalization of strategic decision-focused learning that proves non-monotonic accuracy-payoff curves can make better predictions degrade equilibrium performance, though its algorithmic proposals remain vague, its exogenous-state assumption excludes key interactions, and its real-world benchmarks…
Abstract
Machine learning (ML) predictions are increasingly being used to guide decision-making, giving rise to the problem of decision-focused learning (DFL) where predictors are optimized for downstream decision quality rather than accuracy alone. However, most existing work assumes a single decision-maker optimizing in isolation. This paper formalizes strategic decision-focused learning, where an ML system predicts an exogenous state that some agents observe before playing a game. For example, a park ranger may predict wildlife locations to allocate anti-poaching patrols against strategic poachers. While the exogenous state is unaffected by agent actions, predictions influence agents' strategies and the resulting equilibrium. We find that strategic considerations fundamentally change the learning problem. In particular, we show the prediction accuracy-equilibrium payoff landscape can be non-monotonic, i.e., better predictions can degrade performance. We propose algorithmic approaches to address these challenges and validate them across benchmarks in wildlife conservation and infrastructure protection. Our theory and experiments highlight the importance of accounting for strategic interactions when designing predictors.