The Springer Series on Challenges in Machine Learning

By Frank Hutter et al

Published on June 10, 2019

Read the original document by opening this link in a new tab.

Table of Contents

1. Introduction
2. AutoML Methods
3. AutoML Systems
4. AutoML Challenges

Summary

The document discusses automated machine learning methods, systems, and challenges in the context of the Springer Series on Challenges in Machine Learning. It emphasizes the importance of hyperparameter optimization for improving machine learning algorithms and the reproducibility of scientific studies. The chapter provides an overview of different approaches for hyperparameter optimization, including blackbox optimization algorithms and modern multi-fidelity methods. It highlights the challenges faced in HPO due to the complexity and high dimensionality of the configuration space, as well as the expensive nature of function evaluations for large models.