Abstract
Many today's engineering tasks use approximation of their expensive objective function. Surrogate models, which are frequently used for this purpose, can save significant costs by substituting some of the experimental evaluations or simulations needed to achieve an optimal or near-optimal solution.
This paper presents a surrogate model based on RBF networks. In contrast to the most of the surrogate models in the current literature, it can be directly used for problems with mixed continuous and discrete variables - clustering and generalized linear models are employed for dealing with discrete covariates.
The model has been tested on a benchmark optimization problem and its approximation properties are presented on a real-world application data.