M. Better, F. Glover and M. Laguna
IBM Journal of Research and Development, vol. 51, no. 3/4, pp. 477-487 (2007)
We introduce a simulation optimization approach that is effective in guiding the search for optimal values of input parameters to a simulation model. Our proposed approach, which includes enhanced data mining methodology and state-of-the-art optimization technology, is applicable to settings in which a large amount of data must be analyzed in order to discover relevant relationships. Our approach makes use of optimization technology not only for data mining but also for optimizing the underlying simulation model itself. A market research application embodying agent-based simulation is used to illustrate our proposed approach.
Full text
