Generalization in the Face of Adaptivity: A Bayesian Perspective

Repeated use of a data sample via adaptively chosen queries can rapidly lead to overfitting, wherein the empirical evaluation of queries on the sample significantly deviates from their mean with respect to the underlying data distribution. It turns out that simple noise addition algorithms suffice t...

Full description

Saved in:
Bibliographic Details
Published inarXiv.org
Main Authors Shenfeld, Moshe, Ligett, Katrina
Format Paper
LanguageEnglish
Published Ithaca Cornell University Library, arXiv.org 03.04.2024
Subjects
Algorithms
Data analysis
Queries
Random noise
Variance
Online AccessGet full text

Cover

More Information
Summary:Repeated use of a data sample via adaptively chosen queries can rapidly lead to overfitting, wherein the empirical evaluation of queries on the sample significantly deviates from their mean with respect to the underlying data distribution. It turns out that simple noise addition algorithms suffice to prevent this issue, and differential privacy-based analysis of these algorithms shows that they can handle an asymptotically optimal number of queries. However, differential privacy's worst-case nature entails scaling such noise to the range of the queries even for highly-concentrated queries, or introducing more complex algorithms. In this paper, we prove that straightforward noise-addition algorithms already provide variance-dependent guarantees that also extend to unbounded queries. This improvement stems from a novel characterization that illuminates the core problem of adaptive data analysis. We show that the harm of adaptivity results from the covariance between the new query and a Bayes factor-based measure of how much information about the data sample was encoded in the responses given to past queries. We then leverage this characterization to introduce a new data-dependent stability notion that can bound this covariance.
ISSN:2331-8422