/tabben

A benchmark of datasets for consistently evaluating machine learning models on tabular data.

Primary LanguagePythonMIT LicenseMIT

Open Benchmark on Tabular Data for Machine Learning

This is the main repository for the tabben benchmark for tabular data; it contains the source for the Python package tabben as well as documentation and the scripts used to create the datasets in the benchmark.

  • For a general overview of the project, see our website
  • For things related to the Python package, see the package README or directly with the docs (there are also docs for each included dataset).
  • For the scripts used to generate the datasets, see the scripts README.
  • Otherwise, you may want to see the Contributing Guide for details on how the repo is organized or how to contribute.