auto-sklearn

What auto-sklearn is

auto-sklearn automates fitting machine-learning models and building ensembles for classification and regression. Its Auto-sklearn 2.0 method chooses models, decides whether to use Successive Halving, and draws on portfolios for meta-learning. The fitted pipelines work with the scikit-learn API, making the tool relevant to teams already using that interface. Users can select metrics, define train-validation splits, set feature types, and inspect the search procedure. They can also extend the system by implementing and registering wrapper classes for classification, regression, and feature-preprocessing methods.

The project is free and open source. Its installation documentation covers pip, conda-forge, source, and Docker, and the listed plan requires Linux, Python >=3.7, and a C++11 compiler. auto-sklearn cannot run on Windows because it relies on Python’s Unix-only resource module; the project does not actively support macOS due to resource-limit and dependency issues. Users must set time and memory limits, and runs use one CPU core by default. Parallel workers need access to a shared file system for training data and models. The project is part of the Chair for Machine Learning at the University of Freiburg. It supports model evaluation and API access, and its deployment mode is batch.

Who it suits

auto-sklearn suits machine-learning practitioners who need automated model fitting for classification or regression and work in a Linux environment. It is also a fit for teams that use scikit-learn pipelines and want to customize metrics, splits, feature types, or preprocessing methods.

Limits

The free open-source plan requires Linux, Python >=3.7, and a C++11 compiler. Users must set time and memory limits; the default uses one CPU core, and parallel workers need a shared file system.

Getting started

1. Open the auto-sklearn website and documentation.
2. Choose an installation route: pip, conda-forge, source, or Docker.
3. Use a Linux system with Python >=3.7 and a C++11 compiler.
4. Set time and memory limits for the run.

Verdict

Pick auto-sklearn if you need free, open-source automated fitting for classification or regression and can run it on Linux. Look elsewhere if you need Windows support or macOS support that the project actively maintains.