抄録
In recent years, researchers have become increasingly interested in multi-objective optimization (MOO). However, little work has been done to interpret the MOO solutions in multidimensional space. This paper presents an evolutionary clustering approach to interpret the solutions with its visualization. To comprehend the correlation of each solution and each cluster, the MOO solutions were visualized into the principal component space and the real design space. We recommend that the approach outlined in this study would be applicable to the MOO solutions in the space beyond human perception.
