Research map / Time series / Foundation models for time series
PrimeNet: Pre-training for Irregular Multivariate Time Series
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Pre-train with TimeCL and TimeReco for irregular time series | PrimeNet · PhysioNet | AUC: 0.837 -> 0.842 | Helped |
| Pre-train with TimeCL and TimeReco for irregular time series | PrimeNet · MIMIC-III | AUC: 0.824 -> 0.838 | Helped |
| Pre-train with TimeCL and TimeReco | PrimeNet · Activity at 10% | RMSE: 20.46 -> 14.3 | Helped |
| Pre-train with TimeCL and TimeReco | PrimeNet · PhysioNet at 50% | RMSE: 6.89 -> 3.59 | Helped |
| Pre-train with TimeCL and TimeReco | PrimeNet · Appliances Energy | RMSE: 3.39 -> 3.21 | Helped |
| Add pre-training rather than fine-tuning without pre-training (ablation) | PrimeNet · PhysioNet | AUC: 0.829 -> 0.842 | Helped |
| Add TimeReco to TimeCL (ablation) | PrimeNet · PhysioNet | AUC: 0.811 -> 0.842 | Helped |
| Add TimeCL to TimeReco (ablation) | PrimeNet · PhysioNet | AUC: 0.817 -> 0.842 | Helped |
| Use time-sensitive stratified sampling instead of random sampling (ablation) | PrimeNet with TimeReco · PhysioNet | AUC: 0.834 -> 0.842 | Helped |
| Use constant-time masking instead of constant-length masking (ablation) | PrimeNet with TimeCL · PhysioNet | AUC: 0.829 -> 0.842 | Helped |
Closest papers by meaning
- Irregular Multivariate Time Series Forecasting: A Transformable Patching Graph Neural Networks Approach (ICML 2024 · similarity 0.86)
- Integrating Sequence and Image Modeling in Irregular Medical Time Series Through Self-Supervised Learning (AAAI 2025 · similarity 0.86)
- UP2ME: Univariate Pre-training to Multivariate Fine-tuning as a General-purpose Framework for Multivariate Time Series Analysis (ICML 2024 · similarity 0.85)
- Large Pre-trained time series models for cross-domain Time series analysis tasks (NeurIPS 2024 · similarity 0.85)
- Probabilistic Forecasting of Irregularly Sampled Time Series with Missing Values via Conditional Normalizing Flows (AAAI 2025 · similarity 0.85)
- TimeCHEAT: A Channel Harmony Strategy for Irregularly Sampled Multivariate Time Series Analysis (AAAI 2025 · similarity 0.85)
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.