Advertisement

Model Selection and Error Estimation in a Nutshell

  • Luca Oneto
Book

Part of the Modeling and Optimization in Science and Technologies book series (MOST, volume 15)

Table of contents

  1. Front Matter
    Pages i-xiii
  2. Luca Oneto
    Pages 1-3
  3. Luca Oneto
    Pages 5-11
  4. Luca Oneto
    Pages 13-23
  5. Luca Oneto
    Pages 25-31
  6. Luca Oneto
    Pages 33-57
  7. Luca Oneto
    Pages 59-63
  8. Luca Oneto
    Pages 65-74
  9. Luca Oneto
    Pages 75-86
  10. Luca Oneto
    Pages 87-97
  11. Luca Oneto
    Pages 99-100
  12. Back Matter
    Pages 101-132

About this book

Introduction

How can we select the best performing data-driven model? How can we rigorously estimate its generalization error? Statistical learning theory answers these questions by deriving non-asymptotic bounds on the generalization error of a model or, in other words, by upper bounding the true error of the learned model based just on quantities computed on the available data. However, for a long time, Statistical learning theory has been considered only an abstract theoretical framework, useful for inspiring new learning approaches, but with limited applicability to practical problems. The purpose of this book is to give an intelligible overview of the problems of model selection and error estimation, by focusing on the ideas behind the different statistical learning theory approaches and simplifying most of the technical aspects with the purpose of making them more accessible and usable in practice. The book starts by presenting the seminal works of the 80’s and includes the most recent results. It discusses open problems and outlines future directions for research.

Keywords

Statistical Learning Theory Empirical Data Model Selection Error Estimation Resampling Methods Complexity-Based Methods Union and Shell Bounds Vapnik-Chernovenkis Theory Rademacher Complexity Theory Compression Bound Algorithmic Stability Theory PAC-Bayes Theory Differential Privacy Theory Data-driven models

Authors and affiliations

  • Luca Oneto
    • 1
  1. 1.DIBRISUniversità degli Studi di GenovaGenoaItaly

Bibliographic information