PDF Algebra of Matrices and Vectors 矩阵与向量的代数.pdfVIP

Appendix A Algebr a of Matrice s an d Vector s m o c . c i f i t n e Much of retrieval theory involves solving linear equations , whether th e problem i c s d being solved is linear or non-linear, so a thorough understanding of th e natur e of l r o w . linear equations and th e algebra of matrices an vectors is valuable . Thi s appendix is . y w l n intended t o summarise th e main tools tha t are used in thi s book . For more detailed w o w e s background reading and derivations there many textbook s available at all levels. u m o l r a f n Matrices are denoted by bold face upper case, e.g. A , column vectors by bold d o e s T r d e face lower case, e.g. a, th e transpose by superscript T, e.g. A , so that a row vector a p o r r T - 1 l n o is e.g. b and a vector product is e.g. b a . The inverse is written A and th e F w o 8. transpose of th e inverse may b e written A _ T . D 1 / g 8 n 1 i / d 4 n 0 u o n S o A . l Vecto r Space s c 9 i 4 r . e 4 h 8. p 7 The general linear problem of th e kind encountered in thi s book is of th e form s o 1. m 7 K x = y , where K is an m x n real matrix , with m rows and n columns, x is a t 1 A 1 r y column vector of n unknowns and y is a column vector of m measurements : o b f