Research map / Time series / Causal and generative time series
Probabilistic Time Series Forecasting with Shape and Temporal Diversity
Vincent LE GUEN, Nicolas THOME
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 shape-time diversity sampling to cVAE DILATE (ablation) | cVAE DILATE · Synthetic multimodal | Recall (DILATE): 67.8 ± 7.8 -> 35.5 ± 1.1 | Improved |
| Use DILATE quality training with shape-time diversity instead of MSE quality training (ablation) | DSF shape-time · Synthetic multimodal | Recall (DILATE): 62.0 ± 5.4 -> 35.5 ± 1.1 | Improved |
| Use both shape and time diversity rather than shape diversity alone (ablation) | DSF with DILATE quality · Synthetic multimodal | Recall (MSE): 15.5 ± 1.04 -> 14.5 ± 0.6 | Improved |
| Use both shape and time diversity rather than time diversity alone (ablation) | DSF with DILATE quality · Synthetic multimodal | Recall (MSE): 16.1 ± 0.70 -> 14.5 ± 0.6 | Improved |
| Add shape-time DSF sampling to cVAE DILATE | cVAE DILATE · Traffic | DILATE best (× 100): 17.0 ± 1.1 -> 14.2 ± 0.2 | Improved |
| Add shape-time DSF sampling to cVAE DILATE | cVAE DILATE · Electricity | DILATE best: 22.4 ± 1.3 -> 16.9 ± 0.2 | Improved |
| Add shape-time DSF sampling to cVAE DILATE | cVAE DILATE · M4 | CRPS: 12.2 ± 0.7 -> 10.8 ± 0.5 | Improved |
| Replace the cVAE DILATE baseline with DSF shape-time | Probabilistic forecasters · Traffic | CRPS: 34.4 ± 2.5 -> 29.8 ± 0.3 | Mixed |
Closest papers by meaning
- ProbTS: Benchmarking Point and Distributional Forecasting across Diverse Prediction Horizons (NeurIPS 2024 · similarity 0.86)
- TimeBase: The Power of Minimalism in Efficient Long-term Time Series Forecasting (ICML 2025 · similarity 0.85)
- TimeES: Probabilistic and Deterministic Time Series Forecasting via Evolutionary Spectra (NeurIPS 2026 · similarity 0.85)
- Probabilistic Forecasting of Irregularly Sampled Time Series with Missing Values via Conditional Normalizing Flows (AAAI 2025 · similarity 0.85)
- TimeEmb: A Lightweight Static-Dynamic Disentanglement Framework for Time Series Forecasting (NeurIPS 2025 · similarity 0.84)
- Unlocking the Power of Patch: Patch-Based MLP for Long-Term Time Series Forecasting (AAAI 2025 · similarity 0.84)
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.
