Research map / Time series / Causal and generative time series
Efficient Online Prediction for High-Dimensional Time Series via Joint Tensor Tucker Decomposition
Zhenting Luan, Defeng Sun, Haoning Wang, Liping Zhang
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 larger core tensor (12x4x4 or 80x4x4 vs 3x3x3) (ablation) | TOPA-AAW · USHCN | NRMSE and time: 8.7 vs 5.1 vs 4.8 ms for large, medium, small | Mixed |
| use a longer sliding window (tau=5 to 10 and beyond) (ablation) | TOPA-AAW · USHCN | NRMSE and time: longer window gives better prediction but cost grows with tau | Mixed |
| add automatically adaptive weights (AAW) to TOPA (ablation) | TOPA · Wireless CSI prediction | NRMSE: 0.0636 -> 0.0591 | Improved |
| add AAW to TOPA (ablation) | TOPA · Synthetic TTS | NRMSE: 0.0776 -> 0.0760 | No clear effect |
| add AAW to TOPA (ablation) | TOPA · NASDAQ100 | NRMSE: nearly 20% better than TOPA and MCAR | Improved |
| online one-loop update instead of offline alternating iterations (TOPA vs MCAR) | TOPA · Wireless CSI prediction | time per prediction (ms): 915.8 -> 186.8 | Improved |
| continue online model updating rather than fixed initial model (TOPA-init) (ablation) | TOPA · Synthetic TTS | NRMSE | Improved |
| treat data as tensors rather than vectors (TOPA vs LSTM/GMRL) | TOPA · Synthetic TTS | NRMSE: 0.0776 vs 0.3299 and 0.1390 | Improved |
| use alpha<1 and sufficiently large phi in AAW (ablation) | TOPA-AAW · USHCN and NASDAQ100 | NRMSE | No clear effect |
Closest papers by meaning
- Streaming Factor Trajectory Learning for Temporal Tensor Decomposition (NeurIPS 2023 · similarity 0.86)
- Geometry-aware training of factorized layers in tensor Tucker format (NeurIPS 2024 · similarity 0.84)
- Factor Augmented Tensor-on-Tensor Neural Networks (AAAI 2025 · similarity 0.83)
- Temporal Streaming Batch Principal Component Analysis for Time Series Classification (Student Abstract) (AAAI 2025 · similarity 0.83)
- Functional Complexity-adaptive Temporal Tensor Decomposition (NeurIPS 2025 · similarity 0.83)
- Functional Bayesian Tucker Decomposition for Continuous-indexed Tensor Data (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.
