Research map / Learning theory and algorithms / Matrix and tensor methods
The Decoupled Extended Kalman Filter for Dynamic Exponential-Family Factorization Models
Carlos A. Gomez-Uribe, Brian Karrer
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 block-diagonal per-entity covariance (DEKF) instead of full EKF | EKF · simulated regression, MF, TF (static) | prediction | No clear effect |
| use per-entity blocks instead of full diagonal covariance | DEKF · simulated data | prediction | Improved |
| use full EKF instead of DEKF | DEKF · simulated dynamic tensor factorization | prediction | Improved |
| model parameter dynamics instead of assuming static parameters (ablation) | DEKF · simulated data with drifting parameters | prediction | Improved |
| use Thompson sampling instead of posterior-mean recommendation | DEKF · simulated explore-exploit, static parameters | normalized cumulative regret | Improved |
| increase parameter drift by reducing half life to 10000 then 1000 | Thompson sampling with MF · simulated explore-exploit | cumulative regret | Hurt |
| add parameter dynamics (ablation) | DEKF · MovieLens-20M Gaussian-MF | RMSE | Improved |
| use diagonal DEKF instead of user/item entities | DEKF · MovieLens-20M Bernoulli-MF | normalized cross-entropy | Improved |
| use diagonal DEKF instead of user/item entities | DEKF · MovieLens-20M Gaussian-MF | RMSE | Hurt |
Closest papers by meaning
- Markov Random Fields for Collaborative Filtering (NeurIPS 2019 · similarity 0.86)
- Recommendations with Sparse Comparison Data: Provably Fast Convergence for Nonconvex Matrix Factorization (ICML 2025 · similarity 0.85)
- Influential Exemplar Replay for Incremental Learning in Recommender Systems (AAAI 2024 · similarity 0.85)
- Predictively-Oriented Kalman Filtering (NeurIPS 2026 · similarity 0.84)
- On Efficiency-Effectiveness Trade-off of Diffusion-based Recommenders (NeurIPS 2025 · similarity 0.84)
- Evidential Stochastic Differential Equations for Time-Aware Sequential Recommendation (NeurIPS 2024 · similarity 0.84)
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.
