Parameter Estimation in RSM Taking into Account Errors in Independent Variables

摘要

The regression model taking into account errors in the independent variables is seldom considered in response surface methodology. In response surface design models, independent variables are generally assumed to be measured without errors for given observation. However, in practice the independent variables cannot be exactly controlled and measured, and this situation is widespread in many fields. The aim of this paper is to investigate the regression coefficients, the stationary point and the optimum value in the errors-in-variables model and compare with the model without errors in response surface methodology. Here, as the original model, the independent variables are assumed to be fixed. This result has been obtained using one real data corresponding to a comparison analysis.