Repository for some custom scikit-learn estimators and transformers.

The documentation is viewable here.

• flexible_linear.py implements a scikit-learn estimator for linear regression with custom training and regularization cost functions.
• heavytail.ipynb is a Jupyter notebook demonstrating the use of flexible_linear.py to improve predictions in the presence of heavy-tailed noise. See this post for more details.

This is not a formal package. If anything looks useful to you, please feel free to clone or fork the repository, or just copy individual files.

The documentation is written using Sphinx. To build the documentation locally:

pip install sphinx sphinx_rtd_theme
( cd docs ; make html )

To view the documentation, browse to docs/_build/html/index.html. Alternatively, checkout the gh-pages branch or view the documentation online here.

MIT. See the LICENSE file for more details.