摘要
在设计遗传算法时,主要讨论算法种类的选取和控制参数的设置。基于规则归纳方法,提出一种研究遗传算法种类选择和控制参数设置的试验模型。以遗传算法数值优化问题为背景,用统一结构描述遗传算法试验,通过属性分组等操作规范训练集,利用决策树算法SLIQ进行规则归纳。将获取的规则用于指导遗传算法设计,再用设计出的算法解决某函数的优化求解问题。
In the design of genetic algorithm, the major issues are choosing of algorithm type and parameter setting in genetic algorithm. Based on rule induction, a test model is put forward, which can study the choosing of algorithm type and parameter setting in genetic algorithm. Taking numerical optimization technique using GA as a background, try to give a unified format to describe genetic algorithm experiments, and then regulate the training dataset by some methods, for example, fields grouping. The most important part of this test model is rule induction making use of decision tree algorithm SLIQ. Through that process, get a set of rules which guide the design of genetic algorithm.
出处
《计算机工程与设计》
CSCD
北大核心
2009年第20期4697-4700,4714,共5页
Computer Engineering and Design
基金
国家自然科学基金项目(60743009
60873035)
陕西省自然科学基金项目(2006F43)