@inproceedings{oai:jaxa.repo.nii.ac.jp:00005144,
 author = {中村, 祐輔 and 北村, 健太郎 and 徳光, 政弘 and 石田, 好輝 and 亘, 慎一 and Nakamura, Yusuke and Kitamura, Kentaro and Tokumitsu, Masahiro and Ishida, Yoshiteru and Watari, Shinichi},
 book = {宇宙航空研究開発機構特別資料: 第6回「宇宙環境シンポジウム」講演論文集, JAXA Special Publication: Proceedings of the 6th Spacecraft Enivironment Symposium},
 month = {Feb},
 note = {第6回宇宙環境シンポジウム (2009年2月29日-30日. 北九州国際会議場), 6th Spacecraft Enivironment Symposium (February 29-30, 2009. Kitakyushu International Conference Center), It is important for spacecraft operation to make a forecast of high-energy electron flux at geosynchronous orbit. Because enhancement of high-energy electron flux often causes deep electrical charging. This study to predict a critical enhancement of high-energy electron flux at geosynchronous orbit using the neural network. The program of neural network is tested using the several kinds of input data and various numbers of middle layers. As a result, the prediction efficiency is improved by combining solar wind data and ground magnetic data as input data compared to that by only solar wind data or ground magnetic data. This result indicates that combining solar wind data and ground magnetic data is effective for improvement of the predictive efficiency., 形態: カラー図版あり, Physical characteristics: Original contains color illustrations, 資料番号: AA0064542027, レポート番号: JAXA-SP-09-006},
 publisher = {宇宙航空研究開発機構, Japan Aerospace Exploration Agency (JAXA)},
 title = {ニューラルネットワークによる静止軌道の電子フラックス予測},
 volume = {JAXA-SP-09-006},
 year = {2010}
}