Hyperparameter Optimization: Foundations, Algorithms, Best Practices, and Open Challenges

MCML Authors

Bernd Bischl

Prof. Dr.

Director

Statistical Learning and Data Science

Martin Binder

→ Group Bernd Bischl
 Statistical Learning and Data Science

Michel Lang

Dr.

* Former Member

→ Group Bernd Bischl
 Statistical Learning and Data Science

Tobias Pielok

→ Group Bernd Bischl
 Statistical Learning and Data Science

Janek Thomas

Dr.

* Former Member

→ Group Bernd Bischl
 Statistical Learning and Data Science

Theresa Ullmann

Dr.

* Former Member

→ Group Anne-Laure Boulesteix
 Biometry in Molecular Medicine

Marc Becker

→ Group Bernd Bischl
 Statistical Learning and Data Science

Anne-Laure Boulesteix

Prof. Dr.

Principal Investigator

Biometry in Molecular Medicine

Abstract

Most machine learning algorithms are configured by a set of hyperparameters whose values must be carefully chosen and which often considerably impact performance. To avoid a time-consuming and irreproducible manual process of trial-and-error to find well-performing hyperparameter configurations, various automatic hyperparameter optimization (HPO) methods—for example, based on resampling error estimation for supervised machine learning—can be employed. After introducing HPO from a general perspective, this paper reviews important HPO methods, from simple techniques such as grid or random search to more advanced methods like evolution strategies, Bayesian optimization, Hyperband, and racing. This work gives practical recommendations regarding important choices to be made when conducting HPO, including the HPO algorithms themselves, performance evaluation, how to combine HPO with machine learning pipelines, runtime improvements, and parallelization.

article

Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery

Authors

Links

Research Area

 A1 | Statistical Foundations & Explainability