Research map / Time series / Causal and generative time series
Temporal Regularized Matrix Factorization for High-dimensional Time Series Prediction
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 |
|---|---|---|---|
| TRMF-AR instead of TCF (simple temporal regularizer) | matrix factorization · synthetic forecasting | ND: 1.000 -> 0.373 | Helped |
| TRMF-AR instead of TCF | matrix factorization · traffic forecasting | ND: 0.624 -> 0.187 | Helped |
| TRMF-AR instead of SVD-AR(1) | two-stage SVD plus AR · traffic forecasting | ND: 0.555 -> 0.187 | Helped |
| TRMF-AR instead of DLM | state-space model · walmart-1 forecasting with missing values | ND: 0.602 -> 0.533 | Helped |
| TRMF-AR instead of Mean | baseline predictor · walmart-2 forecasting | ND: 1.097 -> 0.432 | Helped |
| TRMF-AR instead of TCF | imputation · synthetic, 20% observed | ND: 0.713 -> 0.467 | Helped |
| TRMF-AR instead of TCF | imputation · synthetic, 50% observed | ND: 0.289 -> 0.201 | Helped |
| TRMF-AR instead of TCF | imputation · electricity, 20% observed | ND: 0.255 -> 0.245 | No clear effect |
| TRMF-AR instead of DLM | imputation · electricity, 50% observed | ND: 0.158 -> 0.223 | Hurt |
| increase observed fraction from 20% to 50% | TRMF-AR imputation · synthetic | ND: 0.467 -> 0.201 | Helped |
| increase observed fraction from 20% to 50% | TRMF-AR imputation · electricity | ND: 0.245 -> 0.223 | Helped |
| use TRMF-AR rather than AR/DLM at n=50,000 series | TRMF-AR · scalability | runtime: 2 orders of magnitude faster | Helped |
Closest papers by meaning
- TimeEmb: A Lightweight Static-Dynamic Disentanglement Framework for Time Series Forecasting (NeurIPS 2025 · similarity 0.85)
- Structured Matrix Basis for Multivariate Time Series Forecasting with Interpretable Dynamics (NeurIPS 2024 · similarity 0.85)
- Learning Pattern-Specific Experts for Time Series Forecasting Under Patch-level Distribution Shift (NeurIPS 2025 · similarity 0.85)
- HDT: Hierarchical Discrete Transformer for Multivariate Time Series Forecasting (AAAI 2025 · similarity 0.85)
- Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization (ICML 2025 · similarity 0.85)
- Analysing Multi-Task Regression via Random Matrix Theory with Application to Time Series Forecasting (NeurIPS 2024 · 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.