Prior information and eigenvector rotation

Autor:	D. Farrier, D.J. Jeffries
Rok vydání:	2005
Předmět:	Mathematical optimization Covariance matrix Principal component analysis Singular value decomposition MathematicsofComputing_NUMERICALANALYSIS Phase (waves) Singular spectrum analysis Rotation (mathematics) Algorithm Eigenvalues and eigenvectors Matrix decomposition Mathematics
Zdroj:	ICASSP
Popis:	This paper examines a number of methods for including prior information in high resolution eigenvector methods. Both phase and modulus information are considered for this purpose. It is shown that when the phase information contained in rotated eigenvectors is used, the resolution obtained is superior to conventional rotation methods.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::00229fa664f68efb7208721beddaa889 https://doi.org/10.1109/icassp.1987.1169418 Zobrazit plný text záznamu