Learning to route in time and frequency domains: a dual-domain MoE transformer for multi-horizon forecasting

Qi Ji; Jiaxing Wang; Han He; Sheya He; Xiaoyu Dai

doi:10.1038/s41598-026-50232-8

Learning to route in time and frequency domains: a dual-domain MoE transformer for multi-horizon forecasting

Q Qi Ji J Jiaxing Wang H Han He (Institute of Photoelectronic Thin Film Devices and Technology, Renewable Energy Conversion and Storage Center, State Key Laboratory of Photovoltaic Materials and Cells) S Sheya He X Xiaoyu Dai

Download PDF View Original

APA MLA Chicago BibTeX RIS

Article Details

Journal Scientific Reports

Volume / Issue Vol. 16, Issue 1

Published May 11, 2026

DOI 10.1038/s41598-026-50232-8

ISSN 2045-2322

Publisher Nature Portfolio

Journal Info

Scientific Reports

Nature Portfolio

ISSN: 2045-2322 Open Access Life Sciences

Authors (5)

Q

Qi Ji

J

Jiaxing Wang

H

Han He

Institute of Photoelectronic Thin Film Devices and Technology, Renewable Energy Conversion and Storage Center, State Key Laboratory of Photovoltaic Materials and Cells

S

Sheya He

X

Xiaoyu Dai

Learning to route in time and frequency domains: a dual-domain MoE transformer for multi-horizon forecasting

Article Details

Journal Info

Authors (5)

Related Articles from this Journal

AI models of unstable flow exhibit hallucination

The MCAA-YOLO + XPBI integrated model provides a hybrid intelligent measurement method for predicting the body weight of Hu sheep

Application of iron nano-biofertilizers as a sustainable approach to enhance growth and micronutrient uptake in soybean

Adolescents’ status-dependent public and private conformity to prosocial and antisocial decisions in a public goods game

An explainable hierarchical capsule network framework for supporting expert sensory evaluation of herbal medicines with natural language generation