对矩阵进行三角分解，从而生成一个上三角矩阵和下三角矩阵，其乘积等于原矩阵-right triangular matrix decomposition, resulting in the formation of an upper triangular matrix and lower triangular matrices, equivalent to the original product matrix

利用矩阵的奇异值分解压缩图像，先将图像分块，对各块中的每个对应象素内积得到相关矩阵A，然后对A进行奇异值分解，选择n（n<N）个较大的特征值对应的特征向量，利用该向量与各块图像内积来压缩和恢复图像-matrix of singular value decomposition image compression, image first block of the block of each pixel corresponding to be associated with the plot

实现矩阵的Cholesky分解,用Cholesky分解求ARMA模型的参数并作谱估计,利用分裂基算法求复序列 的DFT.将得到的 解卷绕，得到无跳变的相频特性.计算七类窗函数并给出归一化对数幅频曲线-achieve Matrix Cholesky decomposition, Cholesky decomposition used for ARMA model and the parameters for spectral estimation, the use of split-based a

在实现了系数矩阵的三角分解后,为了避免“小主元”做除数而导致更大的误差(In the realization of the coefficient matrix of the triangle decomposition, in order to avoid the "principal" do divisor and cause greater error)