论文部分内容阅读
本文首先建立多层前馈二阶神经网络模型，继而给出该模型的二阶Ｂ－Ｐ学习算法，在此基础上构造了二阶快速Ｂ－Ｐ（即ＦＢ－Ｐ）和改进的二阶ＦＢ－Ｐ（即ＭＦＢ－Ｐ）学习算法，在计算机上以两类飞机图像目标识别为例，对本文提出的多层前馈二阶神经网络模型及其三种二阶学习算法的性能进行仿真实验，并与传统的多层前馈一阶神经网络及其相应学习算法的性能作比较，从而获得若干有意义的结果。
In this paper, a multi-layer feedforward second-order neural network model is established first, and then a second-order B-P learning algorithm for the model is given. Based on this, a second order fast B-P FB-P (MFB-P) learning algorithm is used to simulate the performance of the second-order multi-layer neural network model and its two second-order learning algorithms proposed in this paper. Experiments, and compared with the traditional multi-layer feedforward first-order neural network and its corresponding learning algorithm performance, so as to obtain a number of meaningful results.