The reusable holdout: preserving validity in adaptive data analysis

DOI10.1126/SCIENCE.AAA9375MaRDI QIDQ2961909zbMATH OpenOpenAlexWikidataFDO

Authors Cynthia Dwork, Vitaly Feldman, Moritz Hardt, Toniann Pitassi, Omer Reingold, Aaron Roth

Publication date 15 February 2017

Published in Science (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1126/science.aaa9375

Mathematics Subject Classification ID

Applications of statistics (62Pxx) Foundations and philosophical topics in statistics (62A01)

Recommendations

• Preserving statistical validity in adaptive data analysis (extended abstract)
• Algorithmic stability for adaptive data analysis
• Statistical proof? The problem of irreproducibility
• Algorithmic stability for adaptive data analysis
• Inferactive data analysis

Cited in

(23)

• Sample complexity bounds on differentially private learning via communication complexity
• Test Data Reuse for the Evaluation of Continuously Evolving Classification Algorithms Using the Area under the Receiver Operating Characteristic Curve
• Preserving statistical validity in adaptive data analysis (extended abstract)
• scientific article; zbMATH DE number 7626742 (Why is no real title available?)
• Differential privacy may have a potential optimization effect on some swarm intelligence algorithms besides privacy-preserving
• Separating adaptive streaming from oblivious streaming using the bounded storage model
• A polynomial time, pure differentially private estimator for binary product distributions
• Private PAC learning may be harder than online learning
• scientific article; zbMATH DE number 7370550 (Why is no real title available?)
• Post-selection inference via algorithmic stability
• Algorithmic stability for adaptive data analysis
• Private sequential learning
• Unifying different theories of conformal prediction
• Subsampling suffices for adaptive data analysis
• Comment
• Comment
• Rejoinder
• A Bias-Accuracy-Privacy Trilemma for Statistical Estimation
• Approval policies for modifications to machine learning‐based software as a medical device: A study of bio‐creep
• Discussion on “Approval policies for modifications to machine learning‐based software as a medical device: A study of bio‐creep” by Jean Feng, Scott Emerson, and Noah Simon
• Stability is stable: connections between replicability, privacy, and adaptive generalization
• Making progress based on false discoveries
• Forks over knives: predictive inconsistency in criminal justice algorithmic risk assessment tools

This page was built for publication: The reusable holdout: preserving validity in adaptive data analysis

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2961909)