为了能够提升分解矩阵的稀疏表达能力,提出了一种新的基于平滑ι0范数的正交子空间非负矩阵分解方法。通过将分解矩阵的正交性及平滑ι0范数约束同时引入矩阵分解的目标函数中一起进行优化,大大降低了计算复杂度,并提升了分解矩阵的稀疏表达能力。同时给出了分解矩阵的乘积更新迭代规则。通过在三个真实数据库(Iris,UCI,ORL)上的实验表明,该方法在分解所得矩阵的稀疏表示方面及将其应用于聚类问题所取得的聚类效果方面优于其他方法。
In order to improve the ability of the sparse representations of the NMF, this paper proposed the new algorithm for nonnegative matrix faetorization, denoted smoothed ι0 norm constrained nonnegative matrix faetorization on orthogonal sub- space, in which the generation of orthogonal factor matrices with smoothed ι0 norm constrained were the parts of objective func- tion minimization. Also it developed simple multiplicative updates for the proposed method. Experiments on three real-world databases (Iris, UCI, ORL) show that the proposed method can achieve the best or close to the best in clustering and in the way of the sparse representation than other methods.