AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Temporal Label Smoothing for Early Event Prediction

Hugo Yèche, Alizée Pace, Gunnar Ratsch, Rita Kuznetsova

ICML 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Learning from unlabeled and positive-unlabeled data · 1 citations (OpenAlex)

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
Apply temporal label smoothing instead of multi-horizon predictionEarly event prediction · Circulatory Failure (HiRID)Event Recall: 85.2 $\pm$ 1.7 -> 92.5 $\pm$ 0.5Improved
Apply temporal label smoothing instead of multi-horizon predictionEarly event prediction · Decompensation (MIMIC-III)AUPRC: 34.9 $\pm$ 0.3 -> 35.5 $\pm$ 0.3Improved
Apply temporal label smoothing instead of multi-horizon predictionEarly event prediction · Circulatory Failure (HiRID)Event Recall: 303 to 152 out of 2045 eventsImproved
Use exponential temporal smoothing rather than q_step (ablation)TLS · Circulatory FailureAUPRC: 39.3 $\pm$ 0.2 -> $\mathbf{40.6}$ $\pm$ 0.3Improved
Use exponential temporal smoothing rather than q_shift (ablation)TLS · DecompensationAUPRC: 34.5 $\pm$ 0.4 -> $\mathbf{35.5}$ $\pm$ 0.3Improved
Use exponential temporal smoothing rather than q_linear (ablation)TLS · Circulatory FailureAUPRC: 39.4 $\pm$ 0.3 -> $\mathbf{40.6}$ $\pm$ 0.3Improved
Use exponential temporal smoothing rather than q_sigmoid (ablation)TLS · DecompensationAUPRC: 34.9 $\pm$ 0.4 -> $\mathbf{35.5}$ $\pm$ 0.3Improved
Use exponential temporal smoothing rather than q_concave (ablation)TLS · Circulatory FailureAUPRC: 39.4 $\pm$ 0.3 -> $\mathbf{40.6}$ $\pm$ 0.3Improved
Use multiple horizon-specific outputs rather than shared q_step targets (ablation)Multi-horizon prediction · Circulatory FailureEvent Recall: 83.4 $\pm$ 1.2 -> 85.2 $\pm$ 1.7Improved
Use multiple horizon-specific outputs rather than shared q_step targets (ablation)Multi-horizon prediction · DecompensationEvent Recall: 70.4 $\pm$ 0.7 -> 70.3 $\pm$ 0.6No clear effect
Apply class reweighting rather than ordinary cross-entropyEarly event prediction · Circulatory Failure and DecompensationAUPRCHurt
Replace fixed labels with prediction-dependent temporal targets as in TCSREarly event prediction · Circulatory Failure and DecompensationAUPRCHurt

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