Selective Labels

Follow this topic Ask about this topic

Selective labels arise when decisions affect which outcomes become visible, as when lenders observe repayment only for approved applicants or clinicians observe outcomes shaped by treatment. Researchers use causal identification, propensity weighting, doubly robust estimation, and reject-inference methods to correct for selection, while randomized exploration and audits can collect new outcome data. These methods are applied to fairness, risk evaluation under distribution shift, credit scoring, healthcare, and fraud detection. A central challenge is that results depend on assumptions about unobserved outcomes, and some settings allow only partial identification rather than a single reliable estimate.

Latest papers

10 papers1 paper in the last four weeks, with none the four weeks before. 0.0% of all new papers.

Performative Prediction with Selective Labels
Oct 6, 2026Giovani Valdrighi, Isabel Valera, Marcos Medeiros RaimundoPerformative Prediction Selective Labels
Instituto de Computação Universidade Estadual de Campinas Campinas, Brazil · Department of Computer Science Saarland University Saarbrücken, Germany
Learning Under Treatment-Induced Label Indeterminacy with Expert Annotations of Counterfactual Outcomes: A Case Study in Neurological Prognostication
Aug 12, 2026Xiaobin Shen, Chloe Y. H. Huang, Jonathan Elmer +1Counterfactual Evaluation Clinical Outcome Prediction
Heinz College of Information Systems and Public Policy, Carnegie Mellon University · Department of Emergency Medicine, University of Pittsburgh
When Can Fraud Operations Authorize Automation? A Decision-Support Framework for Fresh Audit Evidence and Review Workload
Aug 9, 2026Jie DengAI-Assisted Decision Making Algorithmic Auditing
Tongji University, Shanghai, China
The Illusion of Improvement: Reject Inference Strategies in Credit Scoring
Jun 16, 2026Bruno Scarone, Ricardo Baeza-YatesCredit Scoring Selection Bias
Northeastern University, Boston, MA, USA · KTH Royal Institute of Technology, Stockholm, Sweden
Identification and Inference for Algorithmic Frontiers with Selective Labels
Jun 12, 2026Yiqi Liu, Francesca Molinari, Amilcar VelezDebiased ML Parameter Identifiability
Department of Economics, Cornell University
Beyond the Training Distribution: Evaluating Predictions Under Distribution Shift and Selection Bias
Jun 12, 2026Annie Ulichney, Amanda CostonDouble ML Distribution Shift Robustness
Department of Statistics, University of California, Berkeley
Post-Rejection Follow-up Sampling: Measuring Outcomes of Rejected Decisions in Algorithmic DEX Trading
Jun 6, 2026Arati Uday KamatCounterfactual Evaluation Decentralized Finance
Independent researcher, United States
Causal Label Recovery in Payment Networks
May 28, 2026Gaurav DhamaFinancial Fraud Detection Learning with Noisy Labels
Mastercard
Long-term Fairness with Selective Labels
May 21, 2026Giovani Valdrighi, Isabel Valera, Marcos Medeiros RaimundoReinforcement Learning Algorithmic Fairness
Instituto de Computac¸˜ao, Universidade Estadual de Campinas, Campinas, Brazil · Department of Computer Science, Saarland University, Saarbrücken, Germany
Confronting Label Indeterminacy in Automated Bail Decisions
Apr 13, 2026Cor Steging, Tadeusz ZbiegieńSelective Labels Learning with Missing Data
Bernoulli Institute of Mathematics, Computer Science and Artificial Intelligence, University of Groningen, The Netherlands · Department of Legal Theory, Jagiellonian University, Poland

Oct 6, 2026J/K move · Enter open · S save

Performative Prediction with Selective Labels

Giovani Valdrighi, Isabel Valera, Marcos Medeiros Raimundo

Instituto de Computação Universidade Estadual de Campinas Campinas, Brazil · Department of Computer Science Saarland University Saarbrücken, Germany

Performative Prediction Selective Labels

Open paper AskPDF arXiv

Many social applications of machine learning exhibit performative effects: population behavior changes in response to deployed models. Performative prediction studies this interaction through a distribution map that relates each model to the population distribution it induces. One of the main results in this framework showed that repeated risk minimization (RRM), which updates models by retraining on the most recent data, can converge to a stable model that minimizes risk on its own induced distribution. However, existing analyses typically assume access to the complete distributions of features and labels after model deployment, ignoring the possibility of selective labels: observing labels only for the accepted subset of the population. In this work, we formalize performative prediction with selective labels and show that retraining only on observed data can misguide the retraining procedure and undermine the guarantees of convergence to a stable solution. We then propose a worst-case objective based on knowledge of a confidence interval on the probability of a positive label. Applying RRM to this objective permits us to remain within a bounded distance to the true stable point. Under a sensitivity assumption on the conditional label distribution, we further show how previously accepted data can tighten these confidence intervals over time. Experiments in a lending application with fairness regularization show that our robust optimization approach closely matches the performance of RRM with complete label access.