CrossLinear is a linear time-series forecasting model with a plug-in 1D convolution that mixes exogenous variables into the target, reporting strong but marginally-supported benchmark results.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
CrossLinear: Plug-and-Play Cross-Correlation Embedding for Time Series Forecasting with Exogenous Variables
CrossLinear is a linear time-series forecasting model with a plug-in 1D convolution that mixes exogenous variables into the target, reporting strong but marginally-supported benchmark results.