poismf

Creates a non-negative low-rank approximate factorization of a sparse counts matrix by maximizing Poisson likelihood with L1/L2 regularization (e.g. for implicit-feedback recommender systems or bag-of-words-based topic modeling) (Cortes, (2018) doi:10.48550/arXiv.1811.01908), which usually leads to very sparse user and item factors (over 90% zero-valued). Similar to hierarchical Poisson factorization (HPF), but follows an optimization-based approach with regularization instead of a hierarchical prior, and is fit through gradient-based methods instead of variational inference.

Last 30 days

This package has been downloaded 294 times in the last 30 days. More than a random curiosity, but not quite a blockbuster. Still, it's gaining traction! The following heatmap shows the distribution of downloads per day. Yesterday, it was downloaded 2 times.

The following line graph shows the downloads per day. You can hover over the graph to see the exact number of downloads per day.

Last 365 days

This package has been downloaded 4,206 times in the last 365 days. Now we’re talking! This work is officially 'heard of in academic circles', just like those wild research papers on synthetic bananas. The day with the most downloads was Feb 20, 2025 with 66 downloads.

The following line graph shows the downloads per day. You can hover over the graph to see the exact number of downloads per day.