AI Paper Atlas

Research map / Time series / Causal and generative time series

ResCP: Reservoir Conformal Prediction for Time Series Forecasting

Roberto Neglia, Andrea Cini, Michael M. Bronstein, Filippo Maria Bianchi

ICLR 2026 · Time series › Causal and generative time series · approach: Probabilistic and conformal forecasting

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add time-dependent decay weights (ablation)ResCP · Exchange (RNN)Winkler score: 0.0269 -> 0.0264Improved
add time-dependent decay weights (ablation)ResCP · Beijing (RNN)Winkler score: 114.85 -> 106.07Improved
add time-dependent decay weights (ablation)ResCP · Solar (RNN)Winkler score: 107.46 -> 104.24Improved
add time-dependent decay weights (ablation)ResCP · ACEA (RNN)Winkler score: 13.41 -> 12.91Improved
use sliding calibration window (ablation)ResCP · Exchange (RNN)Winkler score: 0.0284 -> 0.0264Improved
use sliding calibration window (ablation)ResCP · ACEA (RNN)Winkler score: 14.80 -> 12.91Improved
use sliding calibration window (ablation)ResCP · Beijing (RNN)Winkler score: 106.07 -> 97.99Hurt
use sliding calibration window (ablation)ResCP · Solar (RNN)Winkler score: 104.70 -> 104.24No clear effect
replace learned attention weights (HopCPT) with untrained reservoir similarityResCP · Solar/Beijing/Exchange/ACEAWinkler score: 112.46 -> 104.24Improved
use reservoir-weighted residuals instead of uniform weights (SCP)ResCP · Solar (RNN)Winkler score: 171.41 -> 104.24Improved
train a quantile-regression readout on reservoir states instead of similarity weightingResCQR · ACEA (RNN)Winkler score: 12.91 -> 34.61Hurt
train a GRU quantile model (CoRNN) instead of untrained reservoirCoRNN · Solar (RNN)Winkler score: 104.24 -> 78.42Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy