Research map / Probabilistic and scientific modeling / Variational and generative sampling
Markov Random Fields for Collaborative Filtering
Harald Steck
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 a dense full-rank Gaussian MRF instead of Mult-dae | Collaborative filtering · ML-20M | nDCG@100: 0.419 -> 0.423 | No clear effect |
| Use a dense full-rank Gaussian MRF instead of Mult-dae | Collaborative filtering · Netflix | nDCG@100: 0.380 -> 0.397 | Improved |
| Use a dense full-rank Gaussian MRF instead of Mult-dae | Collaborative filtering · MSD | nDCG@100: 0.313 -> 0.391 | Improved |
| Use sparse set-wise estimation with 0.5% sparsity and r=0.5 instead of the dense solution (ablation) | Gaussian MRF · MSD | training time: 15 min 45 sec -> 2 min 1 sec | Improved |
| Use sparse set-wise estimation with 0.5% sparsity and r=0.5 instead of the dense solution (ablation) | Gaussian MRF · MSD | nDCG@100: 0.391 -> 0.385 | Hurt |
| Increase r from 0 to 0.5 at 0.5% sparsity (ablation) | Sparse Gaussian MRF · MSD | training time: 21 min 12 sec -> 2 min 1 sec | Improved |
| Increase r from 0 to 0.5 at 0.5% sparsity (ablation) | Sparse Gaussian MRF · MSD | nDCG@100: 0.390 -> 0.385 | Hurt |
| Increase r from 0 to 0.5 at 0.1% sparsity (ablation) | Sparse Gaussian MRF · MSD | training time: 3 min 7 sec -> 39 sec | Improved |
| Increase r from 0 to 0.5 at 0.1% sparsity (ablation) | Sparse Gaussian MRF · MSD | nDCG@100: 0.385 -> 0.381 | Hurt |
| Reduce sparsity from 0.5% to 0.1% at r=0 (ablation) | Sparse Gaussian MRF · MSD | nDCG@100: 0.390 -> 0.385 | Hurt |
| Use 0.5% sparsity with r=0.5 instead of 0.1% sparsity with r=0 (ablation) | Sparse Gaussian MRF · MSD | Recall@50: 0.421 -> 0.424 | Improved |
| Use 0.5% sparsity with r=0.5 instead of 0.1% sparsity with r=0 (ablation) | Sparse Gaussian MRF · MSD | training time: 3 min 7 sec -> 2 min 1 sec | Improved |
Closest papers by meaning
- ImplicitSLIM and How it Improves Embedding-based Collaborative Filtering (ICLR 2024 · similarity 0.87)
- Personalized Federated Collaborative Filtering: A Variational AutoEncoder Approach (AAAI 2025 · similarity 0.86)
- Recommendations with Sparse Comparison Data: Provably Fast Convergence for Nonconvex Matrix Factorization (ICML 2025 · similarity 0.86)
- Preference Diffusion for Recommendation (ICLR 2025 · similarity 0.86)
- Learning-Efficient Yet Generalizable Collaborative Filtering for Item Recommendation (ICML 2024 · similarity 0.86)
- From Pairwise to Ranking: Climbing the Ladder to Ideal Collaborative Filtering with Pseudo-Ranking (AAAI 2025 · 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.
