摘要
从粗糙集理论出发,利用决策表局部最小确定性作为阈值来自主控制产生规则,得到一种改进的增量式学习算法。实验表明,在处理大多数数据集时,该算法效率和生成的规则集对样本的正确识别率都较已有的基于粗糙集的增量式算法IRAA有所提高。
An improved incremental learning algorithm for decision tables was presented within the framework of rough sets theory,which could generate rules under the control of threshold automatically.The local minimal certainty of decision table was used as the threshold.Compared with the IRAA algorithm based on rough sets theory,the simulation results show that the new algorithm is more effectively,and the recognition rate is higher in most cases.
出处
《计算机应用》
CSCD
北大核心
2008年第S2期77-79,共3页
journal of Computer Applications
基金
重庆市教委科学技术研究项目(KJ080510)
重庆邮电大学科研基金资助项目(A2004-46)
关键词
粗糙集
增量式学习
知识获取
决策表
rough set
incremental learning
knowledge acquisition
decision table
