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...
保存先:
| 主要な著者: | , , , , , |
|---|---|
| フォーマット: | 論文 |
| 言語: | 英語 |
| 出版事項: |
Taylor & Francis Group
2025-12-01
|
| シリーズ: | Engineering Applications of Computational Fluid Mechanics |
| 主題: | |
| オンライン･アクセス: | https://www.tandfonline.com/doi/10.1080/19942060.2025.2538180 |
| タグ: |
タグ追加
タグなし, このレコードへの初めてのタグを付けませんか!
|