Research map / Time series / Causal and generative time series
SimDiff: Simpler Yet Better Diffusion Model for Time Series Point Forecasting
Hang Ding, Xue Wang, Tian Zhou, Tao Yao
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 Normalization Independence (normalize future targets with own statistics in training) (ablation) | SimDiff · Weather | MSE: 0.328 -> 0.299 | Improved |
| add Normalization Independence (ablation) | SimDiff · NorPool | MSE: 0.555 -> 0.534 | Improved |
| add Normalization Independence (ablation) | SimDiff · Exchange | MSE: 0.019 -> 0.015 | Improved |
| add Normalization Independence (ablation) | SimDiff · Traffic | MSE: 0.389 -> 0.383 | Improved |
| use Median-of-Means ensemble instead of simple averaging (ablation) | SimDiff · Weather | MSE: 0.305 -> 0.299 | Improved |
| use Median-of-Means ensemble instead of a single inference (ablation) | SimDiff · Weather | MSE: 0.317 -> 0.299 | Improved |
| use Median-of-Means instead of simple averaging (ablation) | SimDiff · ETTh1 | MSE: 0.398 -> 0.394 | Improved |
| add RoPE positional embedding (ablation) | SimDiff · NorPool | MSE: 0.582 -> 0.534 | Improved |
| add RoPE positional embedding (ablation) | SimDiff · ETTh1 | MSE: 0.401 -> 0.394 | Improved |
| add cross-channel attention and skip connections (ablation) | SimDiff · design exploration | MSE | Hurt |
| use linear diffusion timestep embedding instead of sinusoidal or activation-added (ablation) | SimDiff · design exploration | MSE | Improved |
| use cosine noise schedule instead of linear or quadratic (ablation) | SimDiff · design exploration | MSE | Improved |
Closest papers by meaning
- DiffPTS: Rethinking Diffusion ELBO for Probabilistic Time Series Forecasting (NeurIPS 2026 · similarity 0.90)
- Non-stationary Diffusion For Probabilistic Time Series Forecasting (ICML 2025 · similarity 0.90)
- Auto-Regressive Moving Diffusion Models for Time Series Forecasting (AAAI 2025 · similarity 0.89)
- Retrieval-Augmented Diffusion Models for Time Series Forecasting (NeurIPS 2024 · similarity 0.88)
- Conditional Diffusion Model with Nonlinear Data Transformation for Time Series Forecasting (ICML 2025 · similarity 0.87)
- ProbTS: Benchmarking Point and Distributional Forecasting across Diverse Prediction Horizons (NeurIPS 2024 · similarity 0.87)
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.
