这些方法都是基于网关联矩阵的满秩分解。

And all of these approaches are derived from the full rank decomposition technique for the incidence matrix.

youdao

通过矩阵的初等变换可实现矩阵的满秩分解和强满秩矩阵的三角分解。

Full rank decomposition of matrix and triangular decomposition of strongly full rank can well be realized by elementary transformation method of matrix.

youdao

非负矩阵分解（NMF）要求分解得到的左矩阵为列满秩，这限制了它在欠定盲分离（UBSS）中的应用。

The decomposed left matrix of Non-negative Matrix Factorization (NMF) is required to be full column rank, which limits of its application to Underdetermined Blind Source Separation (UBSS).

youdao