Research map / Time series / Causal and generative time series
Bayesian Temporal Factorization for Multidimensional Time Series Prediction
Xinyu Chen, Lijun Sun
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 |
|---|---|---|---|
| add VAR temporal prior instead of independent AR factors (BTMF vs BTRMF) (ablation) | Bayesian temporal matrix factorization · Guangzhou traffic speed prediction, horizon 2 | MAPE (%): 11.37 -> 11.01 | Improved |
| add VAR temporal prior instead of independent AR factors (ablation) | Bayesian temporal matrix factorization · London movement speed prediction, horizon 6 | MAPE (%): 18.10 -> 12.01 | Improved |
| add VAR temporal prior instead of independent AR factors (BTTF vs BTRTF) (ablation) | Bayesian temporal tensor factorization · Pacific temperature prediction, horizon 2 | MAPE (%): 16.33 -> 2.85 | Improved |
| add VAR temporal prior instead of independent AR factors (BTTF vs BTRTF) (ablation) | Bayesian temporal tensor factorization · NYC taxi prediction, horizon 2 | MAPE (%): 50.7 -> 53.9 | Mixed |
| use Bayesian BTMF instead of tuned TRMF | BTMF · Hangzhou metro flow prediction, horizon 2 | MAPE (%): 22.5 -> 25.4 | Hurt |
| use BTMF instead of BPMF without temporal modeling | BTMF · Guangzhou traffic speed imputation, 40% random missing | MAPE (%): 9.69 -> 7.61 | Improved |
| use BTMF instead of TRMF | BTMF · Guangzhou traffic speed imputation, 40% random missing | MAPE (%): 7.77 -> 7.61 | Improved |
| use tensor model HaLRTC instead of BTTF | BTTF · Pacific temperature imputation, 40% random missing | MAPE (%): 1.48 -> 0.67 | Hurt |
| use tensor model BATF instead of BTMF | BTMF · Hangzhou metro flow imputation, 40% random missing | MAPE (%): 24.0 -> 19.9 | Hurt |
Closest papers by meaning
- BayOTIDE: Bayesian Online Multivariate Time Series Imputation with Functional Decomposition (ICML 2024 · similarity 0.89)
- Addressing Spatial-Temporal Heterogeneity: General Mixed Time Series Analysis via Latent Continuity Recovery and Alignment (NeurIPS 2024 · similarity 0.86)
- S4M: S4 for multivariate time series forecasting with Missing values (ICLR 2025 · similarity 0.86)
- Graph-based Forecasting with Missing Data through Spatiotemporal Downsampling (ICML 2024 · similarity 0.86)
- RDPI: A Refine Diffusion Probability Generation Method for Spatiotemporal Data Imputation (AAAI 2025 · similarity 0.86)
- TimeMixer++: A General Time Series Pattern Machine for Universal Predictive Analysis (ICLR 2025 · similarity 0.85)
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.
