/BABF

Primary LanguagePythonMIT LicenseMIT

BABF: Bias Aware Probabilistic Boolean Matrix Facortization

This repository is the code for the UAI 2022 paper, Bias Aware Probablistic Boolean Matrix Factorization, which is the first method to derive the Boolean matrix factorization given column- and row-wise bias. Mainly, the difference between Boolean matrix factorization and bias aware boolean matrix factorization could be seen in the following figure.

Running the code

There are two files in this repository. Bias_BIND.py is the main code for BABF methods. run.py is the example code to run BABF. For convenience, run.py also contains methods to simulate binary matrix with added bias and noise.