Repository files navigation

A library forfactorization machines andpolynomial networksfor classification and regression in Python.

Github repository.

Factorization machines and polynomial networks are machine learning modelsthat can capturefeature interaction (co-occurrence) through polynomial terms.Because feature interactions can be very sparse, it's common to uselow rank,factorized representations; this way, we can learn weights even for featureco-occurrences that haven't been observed at training time.

Factorization machines are popular for recommender systems, as they are ageneralization of matrix completion models.

This package provides:

• coordinate descent algorithm for fitting factorization machines of degree 2 or 3,
• coordinate descent algorithm for fitting polynomial networks of arbitrary degree,
• scikit-learn-compatible API,
• Cython implementations for computationally intensive parts.

Binary packages are not yet available.

The development version of polylearn can be installed from its git repository. Inthis case it is assumed that you have a workingC++ compiler.

1. Obtain the sources by:

   git clone https://github.com/scikit-learn-contrib/polylearn.git

or, if git is unavailable,download as a ZIP from GitHub.

2. Install the dependencies:

   # via pippip install numpy scipy scikit-learn nosepip install sklearn-contrib-lightning# via condaconda install numpy scipy scikit-learn noseconda install -c conda-forge sklearn-contrib-lightning

3. Build and install polylearn:

   cd polylearnpython setup.py buildsudo python setup.py install

The solvers implemented are introduced in[1]. Factorization machines are introducedin[2] and polynomial networks in[3].

• Vlad Niculae, 2016-present