摘要
在吸收了人工神经网络解决非线性问题优点的基础上,针对数据挖掘的特点,提出了一种组合BP神经网络的数据挖掘方法。该方法是按照一定方式将n个BP网络组合成一个复杂的BP网络模型,并运用改进的启发式BP算法以实现有用模式的挖掘,能够大大缩短训练时间和提高挖掘精度。
On the basis of the advantages of artificiaI neural networks in nonlinear problems, we presented an approach combining BP neural network data mining, which combine several BP neural networks to form a complicated BP network model in a certain rule while utilizing an improved heuristic BP algorithm to acquire useful mining patterns which shall greatly reduce training time and increase the computing precision.
出处
《洪都科技》
2009年第2期34-42,共9页
Hongdu Science and Technology
关键词
组合神经网络
数据挖掘
combined BP neural network data mining
