Sparse Component Analysis and Applications
Xianchuan Yu, Dan Hu, Jindong Xu · 2014
In this chapter, we discussed the concepts, fundamental model, and development of SCA (sparse component analysis). We systematically reviewed some well-known SCA methods and presented several new methods (including LC-SCA, I-LC-SCA, PC-SCA, PCO-SCA, WL-SCA, and FSCA). Some practical examples of speech signals and remote sensing images were presented toward the end of the chapter. Considering the aspects of Gaussian noise, dependent signals, speech signals, natural images, and remote sensing images, SCA can separate signals more precisely than FastICA.