Sensitivity analysis studies the effect of small changes in the parameters on the objective function value in mathematical programming. These parameters define the objective function and define the constraints for development of the solution to the problem using mathematical programming. The asymptotic normality study of critical points as a consequence of the sensitivity analysis is carried out using mathematical statistics. This paper studies the effect of small perturbations of the regression parameters on the optimum solution of the response surface model and also the asymptotic normality of the critical point is obtained. In response surface analysis, there is a response variable and some input variables and that the input variables can be controlled by the researcher. The response is a function of the input variables. This function is approximated from a polynomial and the success of response surface methodology depends on this approximation. In this study it is assumed that function can be approximated by a polynomial of second order and the unknown parameters can be estimated using regression techniques. (14 refs.)
展开▼
