论文部分内容阅读
结合作者研制的BP神经网络生成器,对BP网络结构和训练参数选取进行了详细的讨论,在克服传统BP网络缺陷方面进行了有益的探索。
Combined with the BP neural network generator developed by the author, the BP network structure and training parameter selection are discussed in detail, and the beneficial exploration is made in overcoming the defects of the traditional BP network.