| 基于最小二乘法的高维目标减少算法 |
| |
| 引用本文： | 梁小鸥,胡蓉.基于最小二乘法的高维目标减少算法[J].人天科学研究,2011(8):39-42. |
| |
| 作者姓名： | 梁小鸥 胡蓉 |
| |
| 作者单位： | 广东工程职业技术学院,广东广州510520 |
| |
| 摘 要： | 多目标进化算法在许多领域有广泛的应用，大部分文献都只针对二维与三维的测试问题，而应用到高维的情况比较少。随着人们开始越来越多地关注高维目标优化问题，目标减少成为研究的热点。从决策者角度考虑冗余目标问题，提出了基于最小二乘法的目标减小算法。该方法将每个目标函数分段拟合为多条直线，然后两两比较各斜率向量确定最冗余目标对，进而确定冗余目标。
|
| 关 键 词： | 高维目标进化 冗余目标减少 最小二乘法 |
Objective Reduction Using the Least Square Method |
| |
| Abstract: | Nowadays, multi-objective evolutionary algorithms have an acceptable performance in many real world problems. However most of the publications consider problems with two or three objectives, and less in many multi-objectives. Researchers begin paying attention to this problem, and objective reduction has recently become a focus. Based on the de- cision makers, this paper proposes a new algorithm, namely objective reduction using the least square method. Our algorithm fits every objective＇s into multi-strip lines, then determines the most redundant objective couples between each two slope vectors, and considers the most redundant objective. |
| |
| Keywords: | Many Multi-Objective Optimizations Nonessential Objective Reduction the Least Square Method |
| 本文献已被 维普 等数据库收录！ |
|
