205 RBFを利用した満足化トレードオフ法の近似解法

抄録

There are a lot pf requirements in design, actually we need to handle them as we would like to draw high performances of those products. I think optimization might be one possibility for that purpose. However, it is not that easy to derive adequate results. Why? And How to deal with that problem. It has been said that Mulit-objective optimization is key technology to that purpose. However, not many researchers are numning deep into it and perform trade-off analysis which is a key technique in MO. One of the reason might be the number of function calls. In this paper, we are going to propose approximate Satisficing method by using Radial Basis Function networks, to overcome this difficulty, and make trade-off analysis as useful hand tool for designers.