Dissecting Performative Prediction: A Comprehensive Survey

2026-02-10

AI Summary

This work addresses the performance degradation of predictive models caused by environmental distribution shifts upon deployment. It presents a systematic review of performative prediction research and introduces a novel classification framework based on the availability of distribution mapping information, distinguishing between two core optimization objectives: performative stability and performative optimality. Through comprehensive literature synthesis, theoretical analysis, and categorization of modeling approaches, the study integrates existing mechanisms for implementing distribution mappings and their associated solution algorithms. Furthermore, it elucidates latent connections between performative prediction and fields such as game theory and causal inference. By offering a unified theoretical perspective and a methodological roadmap, this work aims to foster interdisciplinary collaboration and advance research in performative prediction.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Stochastic Optimization
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

The field of performative prediction had its beginnings in 2020 with the seminal paper"Performative Prediction"by Perdomo et al., which established a novel machine learning setup where the deployment of a predictive model causes a distribution shift in the environment, which in turn causes a mismatch between the distribution expected by the predictive model and the real distribution. This shift is defined by a so-called distribution map. In the half-decade since, a literature has emerged which has, among other things, introduced new solution concepts to the original setup, extended the setup, offered new theoretical analyses, and examined the intersection of performative prediction and other established fields. In this survey, we first lay out the performative prediction setting and explain the different optimization targets: performative stability and performative optimality. We introduce a new way of classifying different performative prediction settings, based on how much information is available about the distribution map. We survey existing implementations of distribution maps and existing methods to address the problem of performative prediction, while examining different ways to categorize them. Finally, we point out known and previously unknown connections that can be drawn to other fields, in the hopes of stimulating future research.

Problem

• performative prediction
• distribution shift
• distribution map
• machine learning
• predictive model

Innovation

• performative prediction
• distribution map
• performative stability
• performative optimality
• machine learning under distribution shift