AI Paper Atlas

Research map / Time series / Long-term forecasting

MSAR: Next-Scale Autoregressive Forecasting for Time Series via Modular Multi-Scale Decoupling

Fanda Fan, Kuiye Ding, Liming Mao, Wang Bingsong, Yao Wang, Xiaorui Wang, Ruijie Jian, Zhipeng Liu, Luqi Gong, Zhenghua Lu, Chunjie Luo, Jianfeng Zhan

NeurIPS 2026 · Time series › Long-term forecasting · approach: Attention for time series forecasting

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes MSAR, a next-scale autoregressive forecasting framework that decouples multi-scale modeling into modules so that multiscale outputs align with multiscale inputs for time series forecasting.

Problem

Most deep time series forecasters use single-scale modeling or apply multiscale processing only to inputs, creating a misalignment between multiscale inputs and single-scale outputs that limits predictive power.

Approach

A modular scale-wise autoregressive framework that generates forecasts coarse-to-fine across scales, so multiscale processing is applied to the outputs as well as the inputs, rather than only on the input side.

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