Research map / Time series / Causal and generative time series
Dish-TS: A General Paradigm for Alleviating Distribution Shift in Time Series Forecasting
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 |
|---|---|---|---|
| Add Dish-TS | N-BEATS · Electricity multivariate, horizon 24 | MSE: 0.041 -> 0.032 | Helped |
| Add Dish-TS | N-BEATS · ETTh1 multivariate, horizon 24 | MSE: 0.797 -> 0.790 | Helped |
| Add Dish-TS | N-BEATS · ETTm2 multivariate, horizon 24 | MSE: 0.643 -> 0.634 | Helped |
| Add Dish-TS | N-BEATS · Weather multivariate, horizon 24 | MSE: 2.557 -> 2.267 | Helped |
| Use Dish-TS instead of RevIN | Autoformer · ETTh1 multivariate, horizon 24 | MSE: 1.245 -> 1.018 | Helped |
| Increase lookback from 48 to 144 (ablation) | N-BEATS + Dish-TS · Electricity, horizon 48 | MSE: 1.2862 -> 0.7309 | Helped |
| Increase lookback from 48 to 144 (ablation) | N-BEATS + Dish-TS · ETTh1, horizon 48 | MSE: 0.5708 -> 0.5202 | Helped |
| Remove prior guidance (set alpha to 0) (ablation) | Dish-TS · Long series forecasting, lengths greater than 168 | prediction error | Hurt |
| Initialize Conet with uniform or average values instead of normal values (ablation) | Dish-TS · ETTh1 and Weather | MSE | Mixed |
Closest papers by meaning
- PAMod: Modeling Cyclical Shifts via Phase-Amplitude Modulation for Non-stationary Time Series Forecasting (NeurIPS 2026 · similarity 0.87)
- Time-Series Forecasting for Out-of-Distribution Generalization Using Invariant Learning (ICML 2024 · similarity 0.86)
- Learning Pattern-Specific Experts for Time Series Forecasting Under Patch-level Distribution Shift (NeurIPS 2025 · similarity 0.86)
- TimeEmb: A Lightweight Static-Dynamic Disentanglement Framework for Time Series Forecasting (NeurIPS 2025 · similarity 0.86)
- Linear Transformers as VAR Models: Aligning Autoregressive Attention Mechanisms with Autoregressive Forecasting (ICML 2025 · similarity 0.86)
- ProbTS: Benchmarking Point and Distributional Forecasting across Diverse Prediction Horizons (NeurIPS 2024 · 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.