Research map / Probabilistic and scientific modeling / Molecules and proteins
Dynamic Tensor Decomposition via Neural Diffusion-Reaction Processes
Zheng Wang, Shikai Fang, Shibo Li, Shandian Zhe
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 diffusion process only (drop reaction process) (ablation) | DEMOTE · CA Weather, CA Traffic, Server Room | nRMSE | Improved |
| use reaction process only (drop diffusion process) (ablation) | DEMOTE · CA Weather, CA Traffic, Server Room | nRMSE | Improved |
| stratified sampling instead of naive sampling (ablation) | DEMOTE · per-epoch running time | speed-up: ~4x to 22x speed-up | Improved |
| learn embedding trajectories (DEMOTE) instead of NONFAT | DEMOTE vs NONFAT · Synthetic two-mode tensor | RMSE: 0.034 -> 0.032 | Improved |
| replace NONFAT with DEMOTE | NONFAT · CA Traffic R=2 | nRMSE: 0.4461 -> 0.3601 | Improved |
| replace NONFAT with DEMOTE | NONFAT · CA Weather R=3 | nRMSE: 0.7460 -> 0.6109 | Improved |
| replace NONFAT with DEMOTE | NONFAT · Server Room R=3 | nRMSE: 0.1407 -> 0.0403 | Improved |
| nonlinear dynamics (NN-DTND) instead of linear dynamics (NN-DTLD) | NN decomposition with discrete-time prior · CA Weather R=7 | nRMSE: 0.7065 -> 0.6543 | Improved |
| nonlinear dynamics (NN-DTND) instead of linear dynamics (NN-DTLD) | NN decomposition with discrete-time prior · Server Room R=7 | nRMSE: 0.4205 -> 0.4205 | No clear effect |
| nonlinear dynamics (CP-DTND) instead of linear (CP-DTLD) | CP decomposition with discrete-time prior · Server Room R=7 | nRMSE: 0.4208 -> 0.0829 | Improved |
| increase embedding dimension R from 2 to 7 | DEMOTE · CA Weather | nRMSE: 0.6327 -> 0.6354 | No clear effect |
| increase embedding dimension R from 2 to 3 | DEMOTE · Server Room | nRMSE: 0.0536 -> 0.0403 | Improved |
Closest papers by meaning
- Learning Mixtures of Linear Dynamical Systems (MoLDS) via Hybrid Tensor–EM Method (ICLR 2026 · similarity 0.88)
- Nonparametric Decomposition of Sparse Tensors (ICML 2021 · similarity 0.87)
- Understanding Deflation Process in Over-parametrized Tensor Decomposition (NeurIPS 2021 · similarity 0.87)
- Tensor Decompositions Meet Control Theory: Learning General Mixtures of Linear Dynamical Systems (ICML 2023 · similarity 0.87)
- Functional Complexity-adaptive Temporal Tensor Decomposition (NeurIPS 2025 · similarity 0.87)
- DiffATS: Diffusion in Aligned Tensor Space (NeurIPS 2026 · similarity 0.86)
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.
