非负矩阵分解的相关讨论

Non-negative matrix factorization discussion

为高效率地处理存放于矩阵中的数据信息,一般采取将矩阵进行分解的方法.不但可将用于描述问题的原始矩阵的维数大大消减,同时也可以对原始矩阵中存放的大量数据进行压缩和概括.新的矩阵分解思想———非负矩阵分解(NMF)算法,即NMF是在矩阵中所有元素均为非负数约束条件之下的矩阵分解方法。

For the efficient handling of the data stored in the matrix information, in general, a method to decompose the matrix. Can be u^ed to describe not only the problem of the original dimension of the matrix substantially reduction, but the original matrix can be stored in large data compression and generalization. The new matrix decomposition ideas - non-negative matrix factorization （NMF） algorithm, that NMF is a matrix whose elements are all non-negative matrix factor- ization under constraint method.