Get Started
Home
Topics
Search
Library
Research questionHow can synthetic medical time-series generation preserve rare-class patterns across temporal scales for downstream prediction?Aggregate distributional and temporal fidelity can conceal clinically informative patterns from rare classes, especially when those patterns occur at different time scales. Synthetic data may therefore appear realistic while offering limited benefit for downstream prediction.
AI
Diffusion Models
Health
Machine Learning
Research Paper
Statistical Machine Learning
Latest papersRecent research connected to this question, newest first.MedFlow: Class-Aware Multi-Scale Generation for Medical Time-Series SynthesisThe source studies MedFlow, a class-aware multi-scale flow-matching generator, on four public datasets spanning EHR, EEG, and ECG signals. Evidence comes from reported downstream prediction experiments and comparisons with diffusion-based baselines; it does not establish performance beyond those datasets and tasks.research paper · Sep 4, 2026
Related questions
How can synthetic time-series generation preserve temporal and frequency structure when training data are scarce?How can wearable foundation models organize representations for targeted longitudinal women’s-health prediction?How can medical concept encoders adapt text-attributed knowledge-graph context to each patient for EHR prediction?How can multimodal time-series forecasting align sparse text with heterogeneous temporal dynamics?