Response surface methodology (RSM): The collection of the statistical and mathematical methods useful for improving, developing, and optimizing processes with significant applications in the design, development and formulation of the new products, as well as in the improvement of the existing product designs. The extensive applications of such type of methodology are in the industrial world particularly in the situations where many input variables potentially influence some performance measure or the quality characteristic of the product or process. The basic purpose of this methodology is to model response based on the group of the experimental factors, and to determine optimal settings of the experimental factors which maximize or minimize the response. Most of the applications include fitting and checking the adequacy of the models of form.
 
 

The vector β and matrix B contain parameters of the model.