论文部分内容阅读
通过对简单一维微分方程的求解,得出了一类人工神经网络模型的解解析表达式。它是由对称阵的特征值与特征向量表达。根据此解析表达式推出了该神经网络模型在特殊情况下的解析表达式。在特殊情况下,特征值与特征向量的神经网络计算已有很多论文论及。最后利用解析表达式分析了该神经网络解的渐近稳定性态。
By solving a simple one-dimensional differential equation, a solution to the analytical expression of a class of artificial neural network models is obtained. It is expressed by the eigenvalues and eigenvectors of a symmetric matrix. According to this analytical expression, the analytical expression of the neural network model under special circumstances is introduced. In special cases, there are many papers on eigenvalue and eigenvector neural network. Finally, the asymptotic stability of the neural network solution is analyzed by using analytical expressions.