Statistical Learning: Stability is Sufficient for Generalization and Necessary and Sufficient for Consistency of Empirical Risk Minimization

Abstract

Solutions of learning problems by Empirical Risk Minimization (ERM) -- and almost-ERM when the minimizer does not exist -- need to be consistent, so that they may be predictive. They also need to be well-posed in the sense of being stable, so that they might be used robustly. We propose a statistical form of leave-one-out stability, called CVEEE(loo) stability. Our main new results are two. We prove that for bounded loss classes CVEEE(loo) stability is (a) sufficient for generalization, that is convergence in probability of the empirical error to the expected error, for any algorithm satisfying it and, (b) necessary and sufficient for generalization and consistency of ERM. Thus CVEEE(loo) stability is a weak form of stability that represents a sufficient condition for generalization for general learning algorithms while subsuming the classical conditions for consistency of ERM. We discuss alternative forms of stability. In particular, we conclude that for ERM a certain form of well-posedness is equivalent to consistency.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 2004
Accession Number
ADA459857

Entities

People

  • Partha Niyogi
  • Ryan Rifkin
  • Sayan Mukherjee
  • Tomaso Poggio

Organizations

  • Massachusetts Institute of Technology

Tags

Communities of Interest

  • C4I

DTIC Thesaurus Topics

  • Algorithms
  • Artificial Intelligence
  • Computer Science
  • Consistency
  • Continuity
  • Contracts
  • Convergence
  • Data Sets
  • Equations
  • Inverse Problems
  • Learning
  • Probability
  • Probability Distributions
  • Random Variables
  • Stability Conditions
  • Standards
  • Supervised Machine Learning

Readers

  • International Relations and European Studies
  • Mathematical Modeling and Probability Theory.
  • Regression Analysis.