Research map / Probabilistic and scientific modeling / Variational and generative sampling
Probabilistic Non-Negative Matrix Factorization and Its Robust Extensions for Topic Modeling
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| replace Frobenius loss with l2,1-norm loss in probabilistic NMF (ablation) | PNMF · Reuters K=100 | perplexity: 1158 -> 1127 | Helped |
| replace l2,1-norm loss with capped l2,1-norm loss (ablation) | PNMFR1 · Reuters K=100 | perplexity: 1127 -> 1103 | Helped |
| replace Frobenius loss with capped l2,1-norm loss (ablation) | PNMF · TDT K=100 | perplexity: 2016 -> 1793 | Helped |
| add probabilistic constraints on factors (PNMF vs NMF) (ablation) | NMF · TDT K=100 | perplexity: 2418 -> 2016 | Helped |
| add probabilistic constraints on factors (PNMF vs NMF) (ablation) | NMF · 20 Newsgroups K=200 | perplexity: 789 -> 792 | Hurt |
| add probabilistic constraints to l2,1-norm NMF (PNMFR1 vs RNMF) (ablation) | RNMF · 20 Newsgroups K=100 | perplexity: 802 -> 778 | Helped |
| use probabilistic NMF instead of LDA | LDA · Reuters K=100 | perplexity: 1328 -> 1158 | Helped |
Closest papers by meaning
- Algorithms for Generalized Topic Modeling (AAAI 2018 · similarity 0.87)
- Provable Algorithms for Inference in Topic Models (ICML 2016 · similarity 0.87)
- Diagnosing and Improving Topic Models by Analyzing Posterior Variability (AAAI 2018 · similarity 0.86)
- Word Co-Occurrence Regularized Non-Negative Matrix Tri-Factorization for Text Data Co-Clustering (AAAI 2018 · similarity 0.86)
- Discovering Discrete Latent Topics with Neural Variational Inference (ICML 2017 · similarity 0.86)
- A Reduction for Efficient LDA Topic Reconstruction (NeurIPS 2018 · similarity 0.86)
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.