TailBooster: Dual-Layer Generative Framework Boosts Extreme Event Prediction

TUDelft · hf · 2026-08-14

TU Delft proposes TailBooster, a dual-layer generative framework with statistical tail extraction and deep autoencoder cleaning to synthesize operationally valid extreme air-transport events, substantially improving extreme-event prediction accuracy.

Original post →

More from Research

Research channel →