TimeRouter: A unified dynamic routing framework for handling missing data in time series forecasting

Qiang Hua

doi:10.1016/j.knosys.2026.115782

TimeRouter: A unified dynamic routing framework for handling missing data in time series forecasting

2026-03-14

0

PRE

AI

DOI：10.1016/j.knosys.2026.115782

Original

Original request for help

Save

Abstract

En 中文

A Unified Dynamic Routing Framework for Handling Missing Data in Time Series Forecasting • A unified framework for handling missing values in multivariate time series. • Quality-driven adaptation of data fusion strategies. • Robust prediction through complementary variable fusion. • Hierarchical space for seamless plug-and-play of expert modules.

Keywords:

TimeSeriesForecasting

MissingDataHandling

DynamicRouting

DataFusion

MultivariateTimeSeries

TimeRouter: A unified dynamic routing framework for handling missing data in time series forecasting

Abstract

Journal

Organization