摘要
为了获取最小决策规则集,当增加新样本时,传统的方法通常需要对决策表中所有数据重新计算,效率欠佳。从可变精度粗集模型理论出发,讨论了新增记录与已有条件属性等价类的关系及对规则集的影响,在此基础上提出了基于可变精度粗集模型的增量式规则获取算法。通过仿真实验表明,这种增量式算法是可行的。
In order to find a minimal set of rules for a decision table, the classical method cannot effectively deal with new instances added to the universe because of recalculation for the overall set of instance. In this paper, first, the relation of the new instances with the existing condition class,and effect on rule sets are studied when a new instance comes. Second, a new incremental learning algorithm based on variable precision rough set model is presented. Finally, the test results proves this algrothm feasible.
出处
《重庆邮电学院学报(自然科学版)》
2005年第6期709-713,共5页
Journal of Chongqing University of Posts and Telecommunications(Natural Sciences Edition)
基金
国家自然科学基金(60373111)
国家科技部攀登--特别支持基金
教育部科学技术研究重点项目(地方01108)
重庆市教育委员会科学技术研究项目(040505
020511)资助
关键词
粗糙集理论
增量式学习
可变精度粗集模型
决策矩阵
rough set theory
incremental learning
variable precision rough set model
decision matrix