Welcome to
NMFLIB is the early version of a lightweight library including different implementations of Non-negative Matrix Factorization, and variants, built using python, numpy, scipy and some components from scikit learn.
The main application intended for this library is clustering.
Currently, the library includes initial implementations of:
- Classical NMF (Lee and Seung)
- Projective NMF (Yuan and Oja)
- Nonnegative Spectral Clustering (Ding et al.)
- Cluster-NMF (Symmetric variant of Convex-NMF) (Ding et al.)
- Sparse-NMF (Kim and Park)
For any questions you can always email me.