AI Paper Atlas

Research map / Learning theory and algorithms / Matrix and tensor methods

Generalized Separable Nonnegative Matrix Factorization

Junjun Pan, Nicolas Gillis

TPAMI 2019 · Learning theory and algorithms › Matrix and tensor methods · approach: Matrix completion and factorization · 40 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use GS-FGM instead of GSPAgeneralized separable NMF · Documents: tr23 (r=6)Relative approximation quality (%): 63.66 -> 70.70Improved
use GS-FGM instead of GSPAgeneralized separable NMF · Faces: Yale (r=38)Relative approximation quality (%): 57.52 -> 68.26Improved
use GS-FGM instead of GSPAgeneralized separable NMF · Faces: CBCL (r=49)Relative approximation quality (%): 80.73 -> 83.10Improved
use generalized separable model instead of separable SPA-RGS-FGM · Faces: CBCL (r=49)Relative approximation quality (%): 83.10 -> 84.57Hurt
use generalized separable model instead of separable SPA-RGS-FGM · Faces: Yale (r=38)Relative approximation quality (%): 62.94 -> 68.26Improved
use GS prior instead of unconstrained NMF at low noiseGS-FGM · Synthetic middle-point with adversarial noiseRelative approximation errorImproved
use GS prior instead of unconstrained NMF at high noiseGS-FGM · Synthetic middle-point with adversarial noiseRelative approximation errorHurt
use GS-FGM instead of NMF/MV-NMF for ground-truth recoveryGS-FGM · Synthetic random GS matricesDistance to ground truthImproved
use GS-FGM instead of GSPA for index recoveryGS-FGM · Synthetic random GS matricesAccuracy: 100% for all epsilon <= 0.483Improved
use generalized separable factorization instead of NMF on real dataGS-FGM · Documents: NG10 (r=10)Relative approximation quality (%): 91.64 -> 92.41Hurt
use GS-FGM instead of GSPA in run timeGS-FGM · Documents (average time)Computation time (s): 0.091 -> 1.249Hurt
use two-sided SPA* heuristic instead of GSPASPA* · Documents: tr45 (r=10)Relative approximation quality (%): 69.08 -> 68.20Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy