LWI-SVD: Low-rank, Windowed, Incremental Singular Value Decompositions on Time-Evolving Data Sets

使用Hermite标准型、SVD分解和满秩分解求矩阵的广义逆，在此基础上求解线性方程组的解。(Hermite's canonical form, SVD decomposition and full rank decomposition are used to find the generalized inverse of the matrix, and then the solution of the linear equation system is solved.)