Optimization basic principles and univariate methods

When the various factors and interactions affecting the results of an experiment have been identified, separate methods are needed to determine the combination of factor levels which will provide the optimum response. It is first necessary to define carefully 

A good optimization method has two qualities. It produces a set of experimental conditions that provide the optimum response, or at least a response that is close to the optimum and it does so with the smallest possible number of trial experimental steps. Speed and convenience will be extremely important, and it is often sufficient to use a method that gets reasonably close to the true optimum in a small number of steps. 

---