Description: matlab源代码,是非负矩阵分解工具箱,包含乘法更新,最小二乘约束,投影梯度法等方法-Usage:
[W,H] = nmf(X,K,alg,maxiter,speak)
W: output matrix
H: output matrix
X: input matrix
K: number of components
alg: algorithm to use
maxiter: maximum number of iterations
speak: print to screen
Algorithms:
mm: Multiplicative updates method using euclidean distance measure.
cjlin: Projected gradient method
prob: Probabilistic non-negative matrix factorization.
als: Alternating least squares.
alsobs: Alternating least squares with optimal brain surgeon.
Demonstrations:
PET: NMF on a PET dataset
Text: NMF used on a three different datasets Email, medical, and CNN.
Algorithms
mm: Multiplicative update method using euclidean distance measure.
Described in Lee and Seung, 2001, Algorithms for Non-negative Matrix Factorization, Advances in Neural Information Processing Systems 13, 556-562. This algorithm is the most commonly used algorithm to solve NMF.
cjlin: Alternative non-negative least squares using projected gradients.
Author: Chih-Je Platform: |
Size: 15746048 |
Author:沈猴子 |
Hits:
Description: IADM_NNLS:不精确的交替方向方法核标准正则化最小二乘问题,可以解决正则项是核范数的公式,例如低秩表示算法-IADM_NNLS: inexact alternating direction method nuclear standard regularized least squares problem, can solve the formula of regularization is nuclear norm, such as low rank said algorithm Platform: |
Size: 77824 |
Author:dulu |
Hits:
Description: In this paper, we present an efficient method for nonnegative matrix factorization based on the alternating nonnegative least squares framework. Our approach adopts a monotone projected Barzilai–Borwein (MPBB) method as an essential subroutine where the step length is determined without line search.(The Lipschitz constant of the gradient is exploited to accelerate convergence. Global convergence of the proposed MPBB method is established. Numerical results are reported to demonstrate the efficiency of our algorithm) Platform: |
Size: 307200 |
Author:song86
|
Hits: