Development of Multi-objective Evolutionary Algorithm Based on Local Fitness Function Landscapes for Combinatorial Optimization Problems
Project/Area Number |
22700158
|
Research Category |
Grant-in-Aid for Young Scientists (B)
|
Allocation Type | Single-year Grants |
Research Field |
Intelligent informatics
|
Research Institution | Kansai University |
Principal Investigator |
HANADA Yoshiko 関西大学, システム理工学部, 助教 (30511711)
|
Project Period (FY) |
2010 – 2011
|
Project Status |
Completed (Fiscal Year 2011)
|
Budget Amount *help |
¥2,730,000 (Direct Cost: ¥2,100,000、Indirect Cost: ¥630,000)
Fiscal Year 2011: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2010: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
|
Keywords | 最適化 / 進化計算 / 多目的最適化 / 組合せ最適化 / 局所探索 / 遺伝的アルゴリズム / 内挿交叉 / 外挿交叉 / 世代交代モデル / 雑音除去フィルタ / 画像処理 |
Research Abstract |
In this study, we developed a widely applicable multi-objective genetic algorithm that can find a set of well-distributed solutions which approximates the entire Pareto front. Our proposed method consists of an interpolation-direct multistep crossover and an extrapolation-direct multistep crossover which consider the local ruggedness in objective functions. We evaluated the effectiveness of the method against the levels of intensity of ruggedness in each objective function through the experiments with multi-objective NK model that explains various intrinsic structures observed in combinatorial problems. In addition we applied our method to the design of noise removal filters of digital images.
|
Report
(3 results)
Research Products
(23 results)