反向神经网络的缺点

However, practical applications are difficult to be satisfied because of the problems of slow learning and the likelihood of being trapped into a local minimum especially when the size of the network is large.

翻译：传统的BP学习速度慢，以及容易进入局部最大值。

These problems are due to the fact that the learning of BP neural network is mechanical and elementary. It does not have the advanced intelligent characteristic to generalize the previous training experience.

翻译：主要原因在于机械性以及基础性，不能形式化之前的学习过程。