集合写真

Workshop on Data Assimilation and Dimensional Reduction

日時:2023.08.24

場所:Kanda Conference Room, Seminar Room 3 (3rd Floor)

言語:英語

趣旨

The primary objective of our research meeting was twofold. First, we aimed to deepen our understanding of Data Assimilation (DA) by connecting and comparing it with relevant fields. This included exploring various aspects of machine and deep learning, such as deep learning techniques, reservoir computing, and (conditional) Variational Autoencoders (VAE).

Second, we sought to extend DA to weather modification. This involved introducing control theory into DA and exploring dimensional reduction methods for Numerical Weather Prediction (NWP) to improve efficiency and accuracy in weather forecasting.

プログラム

TimeSpeakerContent
09:00~Reception
09:15-09:35Shunji Kotsuki (CU)Opening & MS Project
09:35-10:00Craig Bishop (U. Melb.)Bounded Variable DA
10:00-10:25Fumitoshi Kawasaki (CU)Quantum Data Assimilation
10:25-11:05*Daisuke Hotta (MRI)Manifold aspect of DA
11:05-11:30Pascal Oettli (CU)Detection of Separation Scenario
11:30-11:55Fumitoshi Kawasaki (CU)Model Predictive Control
11:55-13:00Group Photo → Lunch
13:00-13:40*Stephen Penny (Sofar Ocean)Reservoir Computing for DA
13:40-14:20*Haruma Ishida (MRI)Conditional VAE
14:20-14:45Daiya Shiojiri (CU)Sparse Sensing Placement (SSP)
14:45-15:00Break
15:00-15:25Mao Ouyang (CU)SSP and Ens. Sens. Analysis
15:25-15:50Craig Bishop (U. Melb.)Ensemble TLM/ADJ
15:50-16:30*Hiroshi Kera (CU)Adversarial Attack
16:30-16:45AllDiscussion
16:45-16:50Shunji Kotsuki (U)Closing
*: 30 min presentation, 10 min discussion†: 20 min presentation, 5 min discussion

参加申し込み

イベントは終了しました。