Lale (https://github.com/IBM/lale) is a Python library for data science with an emphasis on automation, usability, and interoperability. Lalegpl is an extension of Lale with operators from libraries such as Weka and R which have gpl dependencies.
- Slack channel:
#lale-usersin the IBM Research org - Installation instructions
- Python API documentation
- arXiv paper
The following paper has a technical deep-dive:
@Article{arxiv19-lale,
author = "Hirzel, Martin and Kate, Kiran and Shinnar, Avraham and Roy, Subhrajit and Ram, Parikshit",
title = "Type-Driven Automated Learning with {Lale}",
journal = "CoRR",
volume = "abs/1906.03957",
year = 2019,
month = may,
url = "https://arxiv.org/abs/1906.03957" }