人工神经网络训练样本的相容性检测及化简

Consistency and Reduction for the Training Samples of Artificial Neural Network

针对人工神经网络的训练样本庞大繁杂的特点，运用粗糙集理论对有指导的网络训练样本的相容性及其化简进行简单论述，同时对神经网络的特征值提取给出了一种寻找方法．

Simply tells the method of processing the consistency and reduction of the teached trainingsamples by Rough Sets theory, and gives a way of finding the distinction of ANN.