MAR 29 Tue 10:00～11:30 太陽系小天体セミナー / Solar System Minor Body Seminar zoom
MAR 31 Thu 16:00～17:30 SUPER-IRNET seminar on Euclid mission zoom
APR 1 Fri 10:30～12:00 Solar and Space Plasma Seminar zoom
詳細は以下をご覧下さい
3月29日（火）
キャンパス：三鷹
セミナー名：太陽系小天体セミナー
定例・臨時の別：定例
日時：3月29日（火曜日）10時00分～11時30分
場所：zoom
講演者：鈴木文二
世話人の連絡先
名前：渡部潤一
備考：テレビ会議またはスカイプによる参加も可
3月31日（木）
Campus：online
Seminar：SUPER-IRNET seminar on Euclid mission
Regularly Scheduled/Sporadic：Sporadic
Date and time： 31 March 2022, 4pm – 5:30pm
Place：zoom
Speaker, Affiliation, & Title：
Each talk will be 20 min + 10 min.
- Jean-Charles Cuillandre (CEA Saclay)
“The Euclid space survey and the critical complementary ground-based surveys” - Masamune Oguri (Chiba U.)
“The overview of UNIONS” - Stephanie Escoffier (CPPM)
“Selected topics from pre-launch scientific activities in Euclid”
Facilitator
-Name：Michitoshi Yoshida / Yusei Koyama / Takashi Moriya
Comment：
“SUPER-IRNET” (https://ultimate.naoj.org/superirnet/) is a JSPS Core-to-Core Program to promote wide-field NIR survey astronomy in the 2020s, by building a strong research network of scientists in Japan, US, France, Australia, and Taiwan.
4月1日（金）
Campus： Mitaka
Seminar： Solar and Space Plasma Seminar
Regularly Scheduled/Sporadic： Scheduled
Date and time：1 Apr (Fri), 10:30-12:00
Place： zoom
Speaker：Dr. Yusuke Iida
Affiliation：Niigata University
Title：Development of a prediction model for solar flares and sunspot growth based on Deep Learning
I introduce our laboratory’s research on the development of a prediction model for solar flare occurrence and sunspot growth based on Deep Learning. Machine learning, especially Deep learning, which has been developed remarkably in recent one decades, has been used in various scientific fields and also has been actively incorporated in the heliophysics field.
In particular, the prediction of solar flare is a typical problem in which machine learning has a high potential. Since its success in Bobra et al. (2015), a lot of research works try to develop and expand the prediction model for solar flares. On the other hand, there are still several issues that need to be resolved, such as the mysterious behavior of Multi Layer Perceptron (MLP) models to have better prediction accuracy than Convolutional Neural Network (CNN) models, the dependence on rule-based region selection methods, and so on. In the talk, I will introduce our research on these issues. Starting from the basics of deep learning methods, I want to discuss possible future directions of machine learning and solar physics in addition to the research results so far.
Facilitator
-Name：Takayoshi oba