All these methods have advantages and disadvantages. Some choose the points where the probability of improving the response surface is maximized, e.g., MSL, MPI and MEI methods. These three methods rely on the standard error of the GP regression method to force the algorithm to go back and explore regions where the sampled points are sparse. However, for surrogate based optimization problem, it is not necessary to describe the whole response surface. Instead, it is enough just to delineate a region containing the optimum. MIS is a method not focused on building the whole response surface but only the region containing the optimum. Because studies have shown that the MEI method is superior to MSL and MPI methods