Research map / Label-efficient and robust learning / Recognition and long-tail
A Unified Algorithm for One-Cass Structured Matrix Factorization with Side Information
Hsiang-Fu Yu, Hsin-Yuan Huang, Inderjit Dhillon, Chih-Jen Lin
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 changed | Where | Measured | Effect |
|---|---|---|---|
| use weighted square loss on unobserved entries instead of weighted logistic loss (ablation) | LR-wSQ · bibtex | precision@5: 28.43 -> 29.50 | Improved |
| use weighted square loss on unobserved entries instead of weighted logistic loss (ablation) | LR-wSQ · mediamill, delicious | precision@5: 55.33 -> 55.15 (mediamill) | No clear effect |
| use weighted square loss on unobserved entries instead of weighted logistic loss (ablation) | LR-wSQ · bibtex, mediamill, delicious | training time: 85 -> 22 (bibtex); 437 -> 61 (mediamill); 446 -> 23 (delicious) | Improved |
| use Full approach (all unobserved entries negative) instead of Subsampled (ablation) | one-class MF · multi-label datasets | precision@5 | Improved |
| use logistic loss on observed entries instead of square loss (SQ-wSQ) (ablation) | one-class MF · multi-label and recommender datasets | improvement over SQ-SQ | Improved |
| add weighted square loss on unobserved entries (SQ-wSQ vs SQ-SQ) (ablation) | one-class MF · multi-label and recommender datasets | improvement over SQ-SQ | Improved |
| add Nystrom RBF feature expansion to LR-wSQ (ablation) | LR-wSQ · multi-label datasets | precision@5 | No clear effect |
Closest papers by meaning
- Recommendations with Sparse Comparison Data: Provably Fast Convergence for Nonconvex Matrix Factorization (ICML 2025 · similarity 0.87)
- Indexable Probabilistic Matrix Factorization for Maximum Inner Product Search (AAAI 2016 · similarity 0.86)
- Ordinal Non-negative Matrix Factorization for Recommendation (ICML 2020 · similarity 0.85)
- A Hybrid Collaborative Filtering Model with Deep Structure for Recommender Systems (AAAI 2017 · similarity 0.85)
- Confidence-Aware Matrix Factorization for Recommender Systems (AAAI 2018 · similarity 0.84)
- Learning-Efficient Yet Generalizable Collaborative Filtering for Item Recommendation (ICML 2024 · similarity 0.82)
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.
