Max Hopkins ; Daniel M. Kane ; Shachar Lovett ; Gaurav Mahajan - Realizable Learning is All You Need

theoretics:10093 - TheoretiCS, February 6, 2024, Volume 3 - https://doi.org/10.46298/theoretics.24.2
Realizable Learning is All You NeedArticle

Authors: Max Hopkins ; Daniel Kane ; Shachar Lovett ; Gaurav Mahajan

    The equivalence of realizable and agnostic learnability is a fundamental phenomenon in learning theory. With variants ranging from classical settings like PAC learning and regression to recent trends such as adversarially robust learning, it's surprising that we still lack a unified theory; traditional proofs of the equivalence tend to be disparate, and rely on strong model-specific assumptions like uniform convergence and sample compression. In this work, we give the first model-independent framework explaining the equivalence of realizable and agnostic learnability: a three-line blackbox reduction that simplifies, unifies, and extends our understanding across a wide variety of settings. This includes models with no known characterization of learnability such as learning with arbitrary distributional assumptions and more general loss functions, as well as a host of other popular settings such as robust learning, partial learning, fair learning, and the statistical query model. More generally, we argue that the equivalence of realizable and agnostic learning is actually a special case of a broader phenomenon we call property generalization: any desirable property of a learning algorithm (e.g. noise tolerance, privacy, stability) that can be satisfied over finite hypothesis classes extends (possibly in some variation) to any learnable hypothesis class.


    Volume: Volume 3
    Published on: February 6, 2024
    Accepted on: December 11, 2023
    Submitted on: September 28, 2022
    Keywords: Computer Science - Machine Learning,Statistics - Machine Learning,68Q32
    Funding:
      Source : OpenAIRE Graph
    • AF: SMALL: Finding Models of Data and Mathematical Objects; Funder: National Science Foundation; Code: 1909634

    Consultation statistics

    This page has been seen 274 times.
    This article's PDF has been downloaded 243 times.