TiRex-2: Generalizing to Multivariate Data and Streaming
TiRex-2 introduces a groundbreaking model for multivariate time series forecasting.
TiRex-2 is a recurrent xLSTM-based model designed for multivariate forecasting, incorporating both past and future covariates. It overcomes limitations of existing models by maintaining constant cost during streaming and ensuring strict causality for target variables. This innovative approach allows for effective integration of future-known covariates.
This synthesis was produced from its source by AI; there is no human editor or manual review step. How we work