Interactive Tabu Search vs. Interactive Genetic Algorithm

Abstract : We propose an interactive tabu search (ITS) to be used for the development support of a product that fits a human's feeling. Interactive evolutionary computation (IEC) is one of the technologies used in the development support of products that fit a human's feeling using a computer and person undergoing a communication. The interactive generic algorithm (IGA) is generally used in the IEC. A major problem with the use of the IEC is the increased burden on the IEC user to evaluate multiple solution candidates. Using the ITS instead of the IGA may reduce this burden, because the ITS user chooses only his most favorite solution candidate among multiple solution candidates. We performed a comparison of the search performance using simulations with the ITS and IGA. As a result of this simulation, the search performance of the ITS exceeded that of the IGA by a range from 2% to 10%.