对称矩阵特征向量的神经网络求解

On COmputing Eigenvectors of Any Symmetrical Matrix Using Neural Network

对已有的一种用于求解正定矩阵特征向量（特征值重数为１）的人工神经网络进行了研究．通过细致的数学分析，得出用这种网络同样能求出一般对称矩阵最大（最小）特征值对应的特征向量．

A kind of neural network which had been presented to compute eigenvectors of positive-defined symmet-rical matrix whose degrees of eigenvalues must be one was analysed.It can also be used to compute eigenvec-tors corresponding to the maximum(minimum)eigenvalue of any symmetrical matrix.