Investigation of the impact of token embeddings in Transformer-based models on short-term tropical cyclone track and intensity predictions
Tropical cyclones (TCs) are destructive meteorological phenomena, necessitating accurate predictions of TC track and intensity to reduce risks to human life. This study evaluates three Transformer-based models – vanilla Transformer (Transformer), inverted Transformer (iTransformer), and temporal-var...
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Articolo |
| Lingua: | inglese |
| Pubblicazione: |
Taylor & Francis Group
2025-12-01
|
| Serie: | Engineering Applications of Computational Fluid Mechanics |
| Soggetti: | |
| Accesso online: | https://www.tandfonline.com/doi/10.1080/19942060.2025.2538180 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|