Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Multi-way Interacting Regression via Factorization Machines
Mikhail Yurochkin, Xuanlong Nguyen, nikolaos Vasiloglou
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 MiFM₁ instead of Factorization Machine | Retail demand forecasting · Retail demand forecasting | AMAPE: 92.0 -> 92.4 | Hurt |
| Use MiFM₀ instead of Factorization Machine | Retail demand forecasting · Retail demand forecasting | AMAPE: 92.0 -> 92.45 | Hurt |
| Set FFM prior parameter α=0 instead of α=1 (ablation) | MiFM · Gene data, 3-, 4-, 5-way | prediction accuracy: 0.775 -> 0.771 | Hurt |
| Set FFM prior parameter α=0 instead of α=1 (ablation) | MiFM · Gene data, only 5-way | prediction accuracy: 0.649 -> 0.645 | Hurt |
| Set FFM prior parameter α=0 instead of α=1 (ablation) | MiFM · Retail demand forecasting | AMAPE: 92.4 -> 92.45 | No clear effect |
| Vary FFM prior parameter α (ablation) | MiFM · Simulation, interaction recovery across data scenarios | exact interaction recovery: up to 83%; 75% | Mixed |
| Use lower α for binary data and higher α for continuous data (ablation) | MiFM · Simulation, interaction recovery | interaction recovery | Mixed |
| Set FFM prior parameter α=0 to encourage exploration of deeper interactions (ablation) | MiFM · Retail demand forecasting | number of selected 3-way and 4-way interactions: five 3-way and three 4-way; nine 3-way and four 4-way | Improved |
Closest papers by meaning
- The Kernel Interaction Trick: Fast Bayesian Discovery of Pairwise Interactions in High Dimensions (ICML 2019 · similarity 0.87)
- High-Dimensional Interactions Detection with Sparse Principal Hessian Matrix (JMLR 2020 · similarity 0.85)
- Piecewise Constant and Linear Regression Trees: An Optimal Dynamic Programming Approach (ICML 2024 · similarity 0.82)
- Bayesian Additive Distribution Regression (NeurIPS 2026 · similarity 0.82)
- Permutation-Free High-Order Interaction Tests (ICML 2025 · similarity 0.82)
- Exploring the cloud of feature interaction scores in a Rashomon set (ICLR 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.
