Student Seminar Series – December 20, 2007
University of Minnesota
School of Statistics
College
of Liberal Arts

 

Nonparametric Changepoint Model for Phase II Analysis



Qiqi Deng


Thursday, December 20, 2007
10:00 AM,
115 Ford Hall
Minneapolis, East Bank Campus

Refreshments at 9:30 AM
300 Ford Hall

Abstract



 
 
 
 
In statistical process control, the control chart based on changepoint model does not require prior knowledge about the parameters, making it 
an attractive technique. So far, changepoint control charts are only developed under normal assumption. But when the underlying distribution 
is not normal or unclear, it may not be appropriate. In this paper, we propose a nonparametric changepoint model based on Mann-Whitney 
statistic for ongoing Phase II analysis, which has essentially the same computational complexity as the parametric. Its properties under 
Phase I situation are investigated, showing this nonparametric model has very similar behavior as the one for normal. This provides us 
a great guideline for future work in Phase II analysis.