一种改进的基于粗糙集的增量式学习算法

Improved incremental learning algorithm based on rough sets theory

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.