The Comparison of Two Multiple Testing Methods for Outliers Detection in Nonparametric Profile Monitoring

Subject Areas : International Journal of Industrial Mathematics

M.  Rajabi ¹ (Department of Statistics, Science and Research Branch, Islamic Azad University, Tehran, Iran.)
MR.  Faridrohani ² (Department of Statistics, Shahid Beheshti University, Tehran, Iran.)

Received: 2018-03-12 Accepted : 2018-07-18 Published : 2019-08-01

Keywords: Nonparametric Mixed-effects models, Multiple testing, Profile monitoring, Parametric Bootstrap, Local linear Kernel Smoothing,

Abstract :

Multiple testing is one of the methods to analyze such data. We apply multiple testing methods in profile monitoring. Profile monitoring is a relatively new method of quality control, which is best used when data processing is performed using curves or profiles within a given period. In this research, we introduce two multiple testing methods for outlier profile detection. We compare the performance of these two methods by simulations.