pysterior

pysterior is a machine learning library for Python which aims to make Bayesian parametric regression and classification models accessible and easy to use. The library allows users to construct supervised learning models using an intuitive interface similar to that used by scikit-learn.

More documentation is to come - this is still a very new project.

Under the hood, pysterior uses the implementation of the No-U-Turn Sampler (Hoffman and Gelman, 2011) provided by the thoroughly wonderful pymc3. Pymc3 is a requirement to run pysterior; see the pymc3 page to find out how to get the latest version.

Download pysterior on PyPI: https://pypi.python.org/pypi/pysterior

You can install the latest version of this package from PyPI, where it is currently in alpha. The following models are currently supported:

• regression module:
  • (Bayesian) Linear Regression
  • Ridge Regression
  • Lasso Regression
  • Robust Linear regression (Cauchy-distributed noise)
  • Polynomial regression
    
• classification module:
  • Binary Logistic regression
    
  
  

Coming soon:

• k-class Logistic regression
• More non-linear models