RDIT adds residual diffusion and variance calibration on top of a strong point forecaster, achieving best CRPS on seven of eight datasets and lower PICP distance in most settings.
Studies on time series applications in environmental sciences
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
RDIT: Residual-based Diffusion Implicit Models for Probabilistic Time Series Forecasting
RDIT adds residual diffusion and variance calibration on top of a strong point forecaster, achieving best CRPS on seven of eight datasets and lower PICP distance in most settings.