Performance Comparison of Residual Control Charts for a Count Data Based on Ridge Regression

Authors

Shaimaa Mohamed Yassin, Department of Applied Statistics and Econometrics, Faculty of Graduate Studies for Statistical Research, 12613, Cairo University, EgyptFollow
Salah M. Mohamed, Department of Applied Statistics and Econometrics, Faculty of Graduate Studies for Statistical Research, 12613, Cairo University, EgyptFollow

Abstract

Regression models are important and widely used tools to design relationships between pairs of endogenous and exogenous variables. However, they usually suffer from multicollinearity. This study uses residual control charts on count data (Poisson regression), after resolve multicollinearity problems by utilizing the ridge technique. Residual control charts with Poisson regression were introduced by Filho [1], who employed the principal component to treat the multicollinearity problem and, thereafter, prepared a control chart. However, we selected average run length as the metric for the evaluation of the control chart. We used simulated data and application using real data on water quality. The results corresponding to two types of residual values are consistent after being processed by the ridge method, and the joint-charts show the samples to be at out-of-control limits for count data.

Recommended Citation

Mohamed Yassin, Shaimaa and M. Mohamed, Salah (2022) "Performance Comparison of Residual Control Charts for a Count Data Based on Ridge Regression," Information Sciences Letters: Vol. 11 : Iss. 1 , PP -.
Available at: https://digitalcommons.aaru.edu.jo/isl/vol11/iss1/33