Research map / Learning theory and algorithms / Matrix and tensor methods
Holographic Factorization Machines for Recommendation
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 FM inner product with holographic reduced representation (circular convolution/correlation) (ablation) | FM · Netflix | MSE: 0.746 -> 0.719 | Helped |
| replace FM inner product with HRR (ablation) | FM · IMDb | MSE: 3.545 -> 3.429 | Helped |
| replace FM inner product with HRR (ablation) | FM · MovieLens1M | MSE: 1.144 -> 1.134 | Helped |
| replace FM inner product with HRR (ablation) | FM · Kindle Store | MSE: 0.663 -> 0.645 | Helped |
| replace FM inner product with HRR (ablation) | FM · nine datasets, varying k | relative MSE improvement: +0.1% to +7.2% across k | Helped |
| add MLP over flattened memory traces (HFM+) on top of HFM (ablation) | HFM · Netflix | MSE: 0.719 -> 0.663 | Helped |
| add MLP over flattened memory traces (HFM+) on top of HFM (ablation) | HFM · MovieLens1M | MSE: 1.134 -> 1.142 | Hurt |
| HFM+ vs NeuMF | HFM+ · IMDb | MSE: 3.424 -> 3.374 | Helped |
| HFM+ vs NeuMF | HFM+ · MovieLens1M | MSE: 1.142 -> 1.142 | No clear effect |
| larger latent dimension k (ablation) | HFM vs FM · Table 3 datasets | relative improvement over FM: +7.2% (k=4) -> +1.5% (k=16) on IMDb | Hurt |
Closest papers by meaning
- Rethinking Neural vs. Matrix-Factorization Collaborative Filtering: the Theoretical Perspectives (ICML 2021 · similarity 0.87)
- Hgformer: Hyperbolic Graph Transformer for Collaborative Filtering (ICML 2025 · similarity 0.85)
- FACE: A General Framework for Mapping Collaborative Filtering Embeddings into LLM Tokens (NeurIPS 2025 · similarity 0.84)
- Recommendations with Sparse Comparison Data: Provably Fast Convergence for Nonconvex Matrix Factorization (ICML 2025 · similarity 0.84)
- ImplicitSLIM and How it Improves Embedding-based Collaborative Filtering (ICLR 2024 · similarity 0.83)
- Learning-Efficient Yet Generalizable Collaborative Filtering for Item Recommendation (ICML 2024 · 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.