60 search results - page 8 / 12 » Concept Lattice Reduction by Singular Value Decomposition |
PR
2008
13 years 7 months ago
2008
We describe Nonnegative Double Singular Value Decomposition (NNDSVD), a new method designed to enhance the initialization stage of nonnegative matrix factorization (NMF). NNDSVD c...
BMCBI
2008
13 years 7 months ago
2008
Background: Large biological data sets, such as expression profiles, benefit from reduction of random noise. Principal component (PC) analysis has been used for this purpose, but ...
CVPR
2008
IEEE
14 years 9 months ago
2008
IEEE
Tensor based dimensionality reduction has recently been extensively studied for computer vision applications. To our knowledge, however, there exist no rigorous error analysis on ...
ICPR
2008
IEEE
14 years 1 months ago
2008
IEEE
We present a promising analysis on using the pattern of symmetry in the face to increase the accuracy of three-dimensional face recognition. We introduce the concept of the ‘ave...
NIPS
2003
13 years 8 months ago
2003
Spectral methods for nonlinear dimensionality reduction (NLDR) impose a neighborhood graph on point data and compute eigenfunctions of a quadratic form generated from the graph. W...