基于Lanczos算法的对称非负矩阵分解初始化方法

A Lanczos-based Initialization Method for Symmetric Nonnegative Matrix Factorization

为提高对称非负矩阵分解算法的效率,提出了一种基于Lanczos三角化的对称非负矩阵分解初始化方法。该方法可与现有的对称非负矩阵分解算法相结合取得更高的效率.实验表明,现有的对称非负矩阵分解算法与文中提出的初始化方法相结合可以收敛到一个较优解.

Symmetric nonnegative matrix factorization（SNMF） has widely employed in many areas of applications. For improving the efficiency of SNMF algorithms, the authors proposed a novel Lanczostridiagonalization-based initialization method for SNMF, which can be combined with existing SNMF algorithms and achieve higher efficiency. Experiments showed that the SNMF algorithm which combined the proposed initialization method can converges to a better solution.