Global optimization and stochastic methods
John C. M. Nash · 2014
This chapter looks at some tools designed for awkward problems with potentially many local extrema. It presents the ideas by talking of minima. Generally, but not exclusively, R approaches these types of problems stochastically. Guidance to users is limited to the general comment that stochastic methods can have widely varying success depending on the particular problems presented. Several packages are used for stochastic optimization rgp. Package mcga appears to recode the parameters to 8-bit values and finds the minimum of the encoded function. The chapter uses a well-known multimodal test function called the Rastrigin function to illustrate some of these packages that include: GenSA; DEoptim; RcppDE; smco; and soma.