[3G01-03] Analysis Method Development 4 Sensitivity Analysis

[3G03] Development of a robust nuclear data adjustment method to outliers

*Yuhei Fukui¹, Tomohiro Endo¹, Akio Yamamoto¹, Shuhei Maruyama² (1. Nagoya University, 2. JAEA)

Keywords:nuclear data adjustment, outlier, M-estimation, generalized linear least square method, data assimilation, robust estimation

In conventional nuclear data adjustments, the reliability of the adjusted nuclear data is strongly dependent on the reliability of the experimental data used for the nuclear data adjustments. In this study, a robust nuclear data adjustment method to outliners is newly proposed. We confirmed the appropriatenessThe validity of the adjusted nuclear data is confirmed by twin experiments, even under conditions containing inappropriate experimental dataif the experimental data contains outliners.