Revista Colombiana de Estadística
versão impressa ISSN 0120-1751
Several approximations have been recently proposed to optimize processes in which are identified two responses variables, but there is no agreement in the establishment of a compromise solution among the multiples conflicting conditions. We propose a new methodology which takes simultaneously into account the quality of the predictions, the relative variabilities and the cost or preference of each response. An example illustrates the methodology and the results can be replicated using EXCEL SOLVER.
Palavras-chave : Multiobjective analysis; Dual response systems; Response surfaces; Multiple responses; Simultaneous optimization.