2009 Fiscal Year Final Research Report
Importance Sampling in Promising Search Regions by Asymmetrical Normal Distribution Crossover
| Project/Area Number |
19700228
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Sensitivity informatics/Soft computing
|
|---|
| Research Institution | The Institute of Statistical Mathematics |
|---|
Principal Investigator |
SOMEYA Hiroshi The Institute of Statistical Mathematics, モデリング研究系, 助教 (00333518)
|
|---|
| Project Period (FY) |
2007 – 2009
|
| Keywords | 遺伝アルゴリズム / 進化型計算 / 最適化 |
|---|
| Research Abstract |
In function optimization problems, the search performance of real-coded GAs is still unsatisfactory on several objective functions where, for example, the optima are located far from the center region of the search space or the landscapes have ridge structures. In the study, Asymmetrical Normal Distribution Crossover (ANDX) and eXclusive-MGG model (X-MGG) have been proposed for tackling such objective functions. Through experimental and theoretical analyses, we have deepened the understanding of promising search regions in crossover operations.
|
