auto-sklearn vs Weka
auto-sklearn
automl.github.ioWeka
weka.waikato.ac.nzSide by side
| Figure | auto-sklearn | Weka |
|---|---|---|
| Price from | Free | Free |
| Free plan | Yes | Yes |
| Free trial | — | — |
| Open source | — | Yes |
| Windows | — | Yes |
| macOS | — | Yes |
| Linux | Yes | Yes |
| API | Yes | Yes |
| Self-hosted | Yes | Yes |
| Plans published | 1 | 3 |
| Spec score · Predictive Analytics Software | 7.9 (#13) | 8.2 (#10) |
| Extensibility | Users can add classification, regression, and feature-preprocessing methods by implementing and registering wrapper classes. | New classifiers and filters are implemented by deriving classes from Weka classifier and filter base classes, and packages can expose them in the GUI. |
| What it does | auto-sklearn automates machine learning model fitting and ensemble construction for classification and regression tasks. | Weka is a Java machine-learning workbench for data preparation, classification, regression, clustering, association-rule mining and visualization. |
Plans
In their own words
auto-sklearn
auto-sklearn has no write-up on Specifiction yet; its figures are above.
Weka
Weka is open-source machine-learning software for building and evaluating models, with support for forecasting and time-series modeling. It is available free under the GNU General Public License, with Weka 3.8 as the stable release and 3.9 as the development version. The stable branch receives fixes and feature updates intended to preserve compatibility; development builds may introduce breaking changes. Weka provides a Java API, Javadoc for its API and command-line parameters, and can also be used through Jupyter notebooks. Official releases require Java 8 or later. Downloads are listed for Windows, Mac OS, and Linux, with Intel and ARM options. Both branches include a package manager for community extensions, though installing packages requires internet access. The software includes built-in help and a comprehensive manual, with additional assistance through documentation, mailing list archives, forums, and a mailing list. Distributed derivative works must use the GPL; commercial projects needing Weka code in a non-GPL program may be able to obtain a suitable license. Models serialized in version 3.7 are incompatible with 3.8, and the migrator has an exception for RandomForest.