最小二乘回归分析算法，包含特征值与特征向量的提取、训练样本以及最后的识别，采用加权网络中节点强度和权重都是幂率分布的模型。- Least-squares regression analysis algorithm, Contains the eigenvalue and eigenvector extraction, the training sample, and the final recognition, Using weighted model nodes in the network