Optimal Regularization for Performative Learning
Not on the map. Why?
Abstract (arXiv)
In performative learning, the data distribution reacts to the deployed model - for example, because strategic users adapt their features to game it - which creates a more complex dynamic than in classical supervised learning. One should thus not only optimize the model for the current data but also take into account that the model might steer the distribution in a new direction, without knowing the exact nature of the potential shift. We explore how regularization can help cope with performative effects by studying its impact in high-dimensional ridge regression. We show that, while performative effects worsen the test risk in the population setting, they can be beneficial in the over-parameterized regime where the number of features exceeds the number of samples. We show that the optimal regularization scales with the overall strength of the performative effect, making it possible to set the regularization in anticipation of this effect. We illustrate this finding through empirical evaluations of the optimal regularization parameter on both synthetic and real-world datasets.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 42 of its 51 references that are in the map. First posted on arXiv 14 Oct 2025.
Builds on
42 of 51 references.
References · 42
- 2018Reconciling modern machine learning practice and the bias-variance trade-offcited by 919
- 2019Benign Overfitting in Linear Regressioncited by 544
- 2019Surprises in High-Dimensional Ridgeless Least Squares Interpolationcited by 508
- 2020Performative Predictioncited by 318
- 2020Benign overfitting in ridge regressioncited by 152
- 2017Runaway Feedback Loops in Predictive Policingcited by 97
- 2020On the Optimal Weighted ℓ2 Regularization in Overparameterized Linear Regressioncited by 97
- 2021Outside the Echo Chamber: Optimizing the Performative Riskcited by 95
- 2020Performative Prediction in a Stateful Worldcited by 90
- 2019A Model of Double Descent for High-dimensional Binary Linear Classificationcited by 81
- 2015Analysis of classifiers' robustness to adversarial perturbationscited by 76
- 2024Feedback Loops With Language Models Drive In-Context Reward Hackingcited by 70
30 more
- 2020Stochastic optimization with decision-dependent distributionscited by 65
- 2019Transfer Learning for Nonparametric Classification: Minimax Rate and Adaptive Classifiercited by 64
- 2020Asymptotics of Ridge (less) Regression under General Source Conditioncited by 61
- 2022Data Feedback Loops: Model-driven Amplification of Dataset Biasescited by 59
- 2023Performative Prediction: Past and Futurecited by 52
- 2022Dimension free ridge regressioncited by 50
- 2022Multiplayer Performative Prediction: Learning in Decision-Dependent Gamescited by 42
- 2020Provable Benefits of Overparameterization in Model Compression: From Double Descent to Pruning Neural Networkscited by 40
- 2022Regret Minimization with Performative Feedbackcited by 40
- 2023Performative Prediction with Neural Networkscited by 33
- 2021How to Learn when Data Gradually Reacts to Your Modelcited by 29
- 2019The generalization error of max-margin linear classifiers: Benign overfitting and high dimensional asymptotics in the overparametrized regimecited by 27
- 2020Precise High-Dimensional Asymptotics for Quantifying Heterogeneous Transferscited by 26
- 2024Scaling laws for learning with real and surrogate datacited by 26
- 2023Towards a statistical theory of data selection under weak supervisioncited by 25
- 2022Multi-agent Performative Prediction with Greedy Deployment and Consensus Seeking Agentscited by 23
- 2024High-dimensional Analysis of Knowledge Distillation: Weak-to-Strong Generalization and Scaling Lawscited by 21
- 2024Optimal Ridge Regularization for Out-of-Distribution Predictioncited by 18
- 2023Regularization properties of adversarially-trained linear regressioncited by 17
- 2024Generalization error of min-norm interpolators in transfer learningcited by 12
- 2024Minimum-Norm Interpolation Under Covariate Shiftcited by 11
- 2024Optimal Classification under Performative Distribution Shiftcited by 9
- 2025Spurious Correlations in High Dimensional Regression: The Roles of Regularization, Simplicity Bias and Over-Parameterizationcited by 9
- 2024The Role of Learning Algorithms in Collective Actioncited by 8
- 2024Label Noise: Ignorance Is Blisscited by 8
- 2024Learning the Distribution Map in Reverse Causal Performative Predictioncited by 5
- 2025High-dimensional Analysis of Synthetic Data Selectioncited by 5
- 2023Performativity and Prospective Fairnesscited by 4
- 2025On the Impact of Performative Risk Minimization for Binary Random Variablescited by 3
- 2023Constrained Optimization with Decision-Dependent Distributionscited by 2
Built on by
Citing papers per year: 2026: 1.
1 paper.