Flexible sampling of discrete data correlations without the marginal distributions

Alfredo Kalaitzis, Ricardo Silva

Learning the joint dependence of discrete variables is a fundamental problem in machine learning, with many applications including prediction, clustering and dimensionality reducti…