Concept - Non-negative Matrix Factorization (NMF)
What the slides say
- Listed as a dimensionality-reduction algorithm (s.174).
- Named, together with SVD, as a way to find the projection matrix by matrix decomposition in linear algebra (s.177). No formula or example is given for NMF in this chunk.
Added context
NMF factorizes a non-negative matrix into two non-negative factors, which keeps “topics” additive and easier to interpret than SVD’s signed values. This is background, not from the slides.
Related
Concept - Dimensionality Reduction | Concept - Singular Value Decomposition (SVD) | Concept - Latent Semantic Analysis (LSA)