Automatic robust regression analysis of fusion plasma experiment data based on generative modelling

K. Fujii, C. Suzuki, M. Hasuo

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

We have developed a method to learn a robust regression algorithm from the data. We demonstrated that our model trained in this work outperforms other conventional robust analysis methods in terms of the robustness and the accuracy. The model developed and trained in this work was already integrated in the LHD automatic analysis system. When a new Thomson scattering data arrives, our program runs automatically and provides the fitting result based on the trained model. The results are being used by more than 80 other automatic analysis programs in LHD.

Original languageEnglish
Title of host publication45th EPS Conference on Plasma Physics, EPS 2018
EditorsC. Michaut, J. Berndt, M. Mantsinen, S. Coda, G. Lapenta, S. Weber
PublisherEuropean Physical Society (EPS)
Pages609-612
Number of pages4
ISBN (Electronic)9781510868441
StatePublished - 2018
Externally publishedYes
Event45th EPS Conference on Plasma Physics, EPS 2018 - Prague, Czech Republic
Duration: Jul 2 2018Jul 6 2018

Publication series

Name45th EPS Conference on Plasma Physics, EPS 2018
Volume2018-July

Conference

Conference45th EPS Conference on Plasma Physics, EPS 2018
Country/TerritoryCzech Republic
CityPrague
Period07/2/1807/6/18

Fingerprint

Dive into the research topics of 'Automatic robust regression analysis of fusion plasma experiment data based on generative modelling'. Together they form a unique fingerprint.

Cite this