Hierarchical rank-2 nonnegative matrix factorization (HierNMF2) is an unsupervised algorithm for large-scale document clustering and topic modeling. It is about 20 times faster than LDA with comparable quality. HierNMF2 has also been successfully applied in the area of bioinformatics. This Matlab package is developed for the following paper: Da Kuang, Haesun Park, Fast rank-2 nonnegative matrix factorization for hierarchical document clustering, 2013.
Keywords for this software
References in zbMATH (referenced in 4 articles )
Showing results 1 to 4 of 4.
- Du, Rundong; Drake, Barry; Park, Haesun: Hybrid clustering based on content and connection structure using joint nonnegative matrix factorization (2019)
- Tyrtyshnikov, E. E.; Shcherbakova, E. M.: Methods for nonnegative matrix factorization based on low-rank cross approximations (2019)
- Du, Rundong; Kuang, Da; Drake, Barry; Park, Haesun: DC-NMF: nonnegative matrix factorization based on divide-and-conquer for fast clustering and topic modeling (2017)
- Kuang, Da; Yun, Sangwoon; Park, Haesun: SymNMF: nonnegative low-rank approximation of a similarity matrix for graph clustering (2015)