摘要
介绍了BP网络在手写体数字识别中的应用,并在神经网络中引入部分连接结构。实验证明,在神经网络中引入部分连接功能可以大大增强网络的灵活性,可以根据不同的应用来调节网络密度,以达到网络的最高性能。目前识别率为94.61%,实验结果表明,BP神经网络可成功地用于手写数字字符的识别问题。
The application of BP neural network in the handwritten digit recognition and the use of partial link structure in the neural network are introduced. The experiment has proved that the introduction of the partial link function in the neural network can strengthen the flexibility of the network greatly and network density can be adjusted according to different applications to achieve the best performance of the network. Now the system can achieve a recognition ratio of 94. 61%. The experimental result shows that the BP neural network can be used for handwritten alphanumeric character recognition successfully.
出处
《林业劳动安全》
2005年第2期38-42,44,共6页
Forestry Labour Safety
