The generalization performance of a learning method relates to its predictive capability on independent test data. This module deals with assessing how a model you have trained will perform in practice, and how to select a good model. Students will learn to define generalization, why we use validation and test data, the interplay between bias and variance, bootstrapping and cross-validation methods, and theory that gives bounds and intuitions on generalization.

3.1 Introduction to Generalization

3.2 Generalization Basics

3.3 Learning Theory

3.4 Measuring performance

3.5 Generalization in Practice

3.6 Summary

<aside> ⬅️ ‣

</aside>

<aside> ➡️ 3.1 Introduction to Generalization

</aside>

© 2022 — DD1420 Authors