Research map / Learning theory and algorithms / Kernels and regression theory
ERMMA: Expected Risk Minimization for Matrix Approximation-based Recommender Systems
DongSheng Li, Chao Chen, Qin Lv, Li Shang, Stephen Chu, Hongyuan Zha
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 |
|---|---|---|---|
| shrink gradients of a fraction of examples (ERMMA) instead of standard SGD / RSVD | RSVD · MovieLens 10M | RMSE: 0.8256 -> 0.7670 | Improved |
| ERMMA vs SMA, strongest stable baseline | ERMMA · MovieLens 10M | RMSE: 0.7682 -> 0.7670 | Improved |
| ERMMA vs SMA | ERMMA · Netflix | RMSE: 0.8036 -> 0.8018 | Improved |
| shrink gradients by lambda instead of zeroing a fraction (RSVD 2 variant) (ablation) | SGD matrix approximation · MovieLens 10M | test error | Improved |
| randomly set 10% of gradient updates to zero (ablation) | RSVD · MovieLens 10M | test RMSE | No clear effect |
| set s and lambda above 0.5, around (0.8, 0.8) (ablation) | ERMMA · MovieLens 1M | RMSE | Improved |
| ERMMA vs SMA on item ranking | ERMMA · MovieLens 1M N=5 | Average Precision: 0.7672 -> 0.7689 | Improved |
| ERMMA vs SMA on item ranking | ERMMA · MovieLens 10M N=50 | NDCG@10: 0.6856 -> 0.6976 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- LLORMA — LLORMA: Local Low-Rank Matrix Approximation (JMLR 2016) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Confidence-Aware Matrix Factorization for Recommender Systems (AAAI 2018 · similarity 0.86)
- Low-Rank Matrix Approximation with Stability (ICML 2016 · similarity 0.85)
- LLORMA: Local Low-Rank Matrix Approximation (JMLR 2016 · similarity 0.85)
- GLOMA: Embedding Global Information in Local Matrix Approximation Models for Collaborative Filtering (AAAI 2017 · similarity 0.85)
- Generalization Error Bounds for Two-stage Recommender Systems with Tree Structure (NeurIPS 2024 · similarity 0.84)
- Recommendations with Sparse Comparison Data: Provably Fast Convergence for Nonconvex Matrix Factorization (ICML 2025 · similarity 0.83)
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.
