Optimization problems are abundant in analytical chemistry, while the field of computational science is bringing new optimization methods. These methods differ in strategy, sufficient for some to work very well on one optimization problem, while others are costly, inefficient, or fail at all at finding the optimum. From the chemist's point of view, it can be said that for each optimization problem there exists a suitable method to find the optimum or an acceptable solution.
For problems that can be calculated with a differentiable formula, a method that uses the derivative can find the optimum probably very fast. Problems with multiple local optima are best solved with global optimizers, while for simple problems with only one optimum, a local method is more efficient. Some problems are noisy in the solution, others are too complex to optimize and are better tackled by just trying to find a really good solution instead of the best. This article provides an overview of the established classical and new nonclassical and experimental optimization techniques and highlights the mechanics of the techniques to give an understanding of which technique can be used best for the optimization problem at hand.