• :

     

    In typical analyses of data better-modeled by a nonlinear mixed model, an unusual observation (profile), either within a cluster, or an entire cluster itself, can greatly distort parameter estimates and subsequent standard errors. Consequently, inferences about the parameters are misleading.  This research introduces a novel semiparametric control chart for modeling non-linear profile data, a nonparametric (NP) and a semiparametric procedure that combines both parametric and NP profile fits based on the residuals.  In addition, however, the semiparametric method is robust to model misspecification because it also performs well when compared to a correctly specified nonlinear mixed parametric model.  The proposed control charts showed excellent capability for detecting changes in Phase I data.  An example is given using the mixed parametric logistic model and medical data, comparing the robust approaches to the non-robust one.

     

  • : Abdel-Salam Gomaa Abdel-Salam
  • : Department of Mathematics, Statistics and Physics, Qatar University, Doha, Qatar.
  • : Abdel-Salam Gomaa Abdel-Salam
  • : quality
  • : advanced/theoretical
  • : abdo@qu.edu.qa
Novel Techniques for Non-linear Profile Monitoring via Residuals