论文部分内容阅读
目前,在数十种神经网络模型中以BP算法作为学习方式的多层神经网络是使用得最为广泛的一种。该网络是由Rumechere和Mccelland领导的研究小组在1985年实现的,已有效地解决了XOR、T-C匹配等感知器所不能解决的问题。现已广泛应用于语音识别、模式分类、过程监控等领域。尽管多层神经网络的BP算法较有效,但由于它用于高
At present, the multi-layer neural network which uses BP algorithm as a learning method in dozens of neural network models is the most widely used one. The network was implemented in 1985 by a research team led by Rumechere and Mccelland and has effectively addressed issues that can not be solved by perceptrons such as XOR and T-C matching. Has been widely used in speech recognition, pattern classification, process monitoring and other fields. Although the BP neural network algorithm more effective, but because it is used for high