Differential Evolution, is a recently invented global optimization algorithm. Originally proposed as a method for the global continuous optimization Differential Evolution has been easily modi.ed for handling mixed (continuous and discrete) variables. In order to have a better choice of the Differentiation?s formula, we introduce a generalization of the Differential Evolution?s strategies. This is done by dividing them into four groups according to their differentiation principle. Such approach leads us to the new universal formula of Differentiation. Some examples of strategies are demonstrated and compared on the De Jong test functions.