| Project/Area Number |
15K00338
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Soft computing
|
|---|
| Research Institution | Osaka University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2015-04-01 – 2020-03-31
|
| Project Status |
Completed (Fiscal Year 2019)
|
| Budget Amount *help |
¥4,810,000 (Direct Cost: ¥3,700,000、Indirect Cost: ¥1,110,000)
Fiscal Year 2018: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2017: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2016: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2015: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
|
|---|
| Keywords | 進化計算 / 分布推定アルゴリズム / 反応拡散方程式 / 数理モデル / 群知能 / ブラックボックス関数最適化 / 関数最適化 / 粒子群最適化 / ホタルのアルゴリズム |
|---|
| Outline of Final Research Achievements |
The balance between exploitation and exploration, a key concept in the design of efficient evolutionary computation algorithms is related to both competitive and cooperative behavior. The nature of a system with such conflicting actions is inherently complex, but the reaction-diffusion model has been used to explain it. Although the actual behaviors of evolutionary computation and swarm intelligence are similarly complex, in this study, we aim to give an explanation by using a mathematical model based on competition and cooperation, in other words, repulsion, and chemotaxis. We have shown that a distribution estimation type evolutionary process is realized based on a partial differential equation model having a good property for optimization.
Then, we would give an explanation for an actual behavior of evolutionary computation and swarm intelligence based on the mathematical model. Besides, we have proposed a hybrid type of swarm intelligence model based on such an explanation.
|
| Academic Significance and Societal Importance of the Research Achievements |
進化計算は有用な最適化ツールとして知られているが，その理論的基盤は乏しい．個体の振る舞いを記述するモデルに一般性がないことがその遠因である．
そこで，本課題では，多峰性関数の最適化にとって有用な性質を有する数理モデルに着目し，モデルを通じた進化計算の手続きの説明のために，数理モデルから進化型の手続きを構築した．得られたアルゴリズムでは，効率的な進化計算のアルゴリズムをデザインするうえで重要な概念である Exploitation と Exploration のバランスが陽に説明でき，進化計算をデザインする上で有用な知見が得られた．
|
