Multi-Objective Automated Co-evolution
Co-evolutionary algorithms are a nature inspired approach to problems for which no function for evaluating potential solutions are known. Applying co-evolutionary search to the problem of multi-objective optimisation has a wide range of practical applications; of particular interest is the applic...
Saved in:
| 主要作者: | |
|---|---|
| 其他作者: | |
| 格式: | Final Year Project |
| 語言: | English |
| 出版: |
2011
|
| 主題: | |
| 在線閱讀: | http://hdl.handle.net/10356/44849 |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 機構: | Nanyang Technological University |
| 語言: | English |
| 總結: | Co-evolutionary algorithms are a nature inspired approach to problems for which no
function for evaluating potential solutions are known. Applying co-evolutionary
search to the problem of multi-objective optimisation has a wide range of practical
applications; of particular interest is the application to automated red teaming. At
current however, little research has been done in this field of multi-objective coevolution.
This project attempts to add to the literature by surveying current research on singlesided
co-evolution and from there developing a framework to support multi-objective
competitive co-evolution.
Here, a preliminary design of such a framework is implemented and applied to the
search of optimal strategies for tic-tac-toe. The dynamics of co-evolution and
pathologies existing are thusly analysed. |
|---|