Auto-Sklearn 2.0: Hands-free AutoML via Meta-Learning

TL;DR

Auto-Sklearn 2.0 introduces a fully automated AutoML system that leverages meta-learning and bandit strategies to optimize performance on large datasets within strict time limits, making AutoML more accessible and efficient.

Contribution

The paper presents Auto-sklearn 2.0, a hands-free AutoML system that automates pipeline design and hyperparameter tuning using meta-learning and a novel design space exploration.

Findings

Reduces relative error by up to 4.5 times compared to previous AutoML systems.

Achieves better performance within 10 minutes than Auto-sklearn 1.0 does in an hour.

Demonstrates effectiveness on 39 benchmark datasets.

Abstract

Automated Machine Learning (AutoML) supports practitioners and researchers with the tedious task of designing machine learning pipelines and has recently achieved substantial success. In this paper, we introduce new AutoML approaches motivated by our winning submission to the second ChaLearn AutoML challenge. We develop PoSH Auto-sklearn, which enables AutoML systems to work well on large datasets under rigid time limits by using a new, simple and meta-feature-free meta-learning technique and by employing a successful bandit strategy for budget allocation. However, PoSH Auto-sklearn introduces even more ways of running AutoML and might make it harder for users to set it up correctly. Therefore, we also go one step further and study the design space of AutoML itself, proposing a solution towards truly hands-free AutoML. Together, these changes give rise to the next generation of our AutoML system, Auto-sklearn 2.0. We verify the improvements by these additions in an extensive experimental study on 39 AutoML benchmark datasets. We conclude the paper by comparing to other popular AutoML frameworks and Auto-sklearn 1.0, reducing the relative error by up to a factor of 4.5, and yielding a performance in 10 minutes that is substantially better than what Auto-sklearn 1.0 achieves within an hour.