Monitoring Profiles Based on Simultaneous Confidence Bands for Penalized Spline Regression
Longcheen Huwang
Institute of Statistics, National Tsing Hua University, Hsinchu, Taiwan

In this talk, a penalized spline regression model is used to describe the complex relationship (profile) between a response variable and an explanatory variable. By treating the penalized spline regression model as a linear mixed model, we can estimate the parameters of the penalized spline regression using the method for fitting the linear mixed model. A Shewhart-type simultaneous confidence band (SCB) chart for the penalized spline regression model is proposed for monitoring profiles in Phase II study. Simulation studies are conducted to evaluate the performance of the SCB chart. An industrial example used to demonstrate the applicability of the proposed chart is provided as well.

Keywords: Profile monitoring; Control chart; Average run length; Simultaneous confidence band

Biography: Professor Longcheen Huwang received his PhD in Statistics from Cornell University in 1991. He is currently a Full Professor at the Institute of Statistics, National Tsing Hua University, Taiwan. His research interests include Statistical Inference and Quality Management, especially in Statistical Process Control.