AI Paper Atlas

Research map / Time series / Long-term forecasting

Anti-Aliasing Matters: A Dynamic Network for Time Series Forecasting

Heng Zhou, Xin Sun, Chao Li

ICML 2026 · Time series › Long-term forecasting · approach: Graph-based multivariate forecasting

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 changedWhereMeasuredEffect
replace depthwise+pointwise conv downsampling with linear layers (ablation)DMANet · UnempMSE: 0.064 -> 0.081Hurt
use standard strided convolution instead of depthwise+pointwise (ablation)DMANet · UnempMSE: 0.064 -> 0.075Hurt
use frequency-domain upsampling with channel expansion before zero-padding (ablation)DMANet · UnempMSE: 0.077 (Linear Up) -> 0.064Improved
add RevIN normalization (ablation)DMANet · NASDAQMSE: 2.337 -> 0.177Improved
use frequency-domain MAE loss instead of MSE loss (ablation)DMANet · NASDAQMSE: 0.195 -> 0.177Improved
add pre-sampling anti-aliasing filter (ablation)DMANet generic baseline · UnempMSE: 0.075 -> 0.064Improved
add post-sampling frequency interpolation (ablation)DMANet generic baseline · UnempMSE: 0.069 -> 0.064Improved
use ESR-based dynamic low-pass mask instead of Max/Random/classical filters (ablation)DMANet · PEMS08MSE: 0.117 (Max) -> 0.090Improved
plug ESR-based anti-aliasing filter into TimeMixer and MICNTimeMixer, MICN · multiple datasetsMSE/MAEImproved
add anti-aliasing filter on a high-frequency injected ETTh1 signalDMANet · ETTh1 spectral injectionHigh-Frequency Capture: 99.8% HFC vs 76.9% FilterNetImproved
anti-aliasing filter overheadDMANet · synthetic datalatency: worst-case increase of only 2.4%No clear effect
DMANet vs iTransformer on high-dimensional Solar-EnergyDMANet · Solar-Energy (L=96)MSE: 0.227 vs TimeMixer 0.216Mixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy