摘要
根据数据挖掘中粗糙集理论和BP神经网络各自的优势和存在的问题,提出了一种将粗糙集理论和BP神经网络理论结合在一起的算法。该算法利用粗糙集对属性的归约功能将数据仓库中的数据进行归约,并将归约后的数据作为训练数据提供给BP神经网络。通过粗糙集归约,提高了训练数据表达的清晰度,也减小了BP神经网络的规模,同时利用BP神经网络又克服了粗糙集对噪声数据敏感的影响。文中提出了代价函数,解决了训练数据与网络精度的问题,也提供了由粗糙集归约向BP神经网络训练转变的依据。
According to the advantages and the problems exiting in rough sets theory and neural network of data mining,an algorithm is presented based on the combination of rough sets theory and BP neural network.This algorithm reducts data from data warehouse by using rough sets reduct function,and then transfers the reducted data to the BP neural network as training data.By data reduct,the expression of training will become clear,and the scale of neural network can be simplified.At the same time ,neural network can solve rough sets' problem of yawp sensitivity.This paper presents a cost function to express the relationship between the amount of training data and the precision of neural network,and to supply the standard for the change from rough set reduct to neural network training.
出处
《计算机工程与应用》
CSCD
北大核心
2004年第31期169-172,175,共5页
Computer Engineering and Applications
关键词
数据挖掘
粗糙集
神经网络
data mining,rough sets,neural network