Sinwook Lee Digital Media Lab. HYU
Linear Equations To gain the appropriate solution, 1..n of equations are necessary. The num of equations < n underdetermined there may be many solutions The num of equations > n overdeterminded there may be no solutions
Letting A = (a ij ), x = (x i ), b = (b i ), as Ax = b A is nonsingular
“LUP decomposition” is introduced for a fast and stable method The process with LUP decomposition The process to fine LUP matrices effeciently
The idea behind LUP is to find three n x n matrices L, U, and P such that PA = LU where L is a unit lower-triangular matrix U is an upper-triangular matrix P is a permutation matrix
LUP decomposition satisfied for Ax = b, PA = LU PAx = Pb LUx = Pb, Let us define y = Ux, where x is the desired solution vector Ly = Pb In the Ly = Pb, we solve the unknown vector y by a method called “forward substitution” In the y = Ux, we solve the unknown vector x by a method called “backward substitution”
….. Ly = Pb
Ux = y
PA = LU We start with the case A is an nxn nonsingular matrix P is absent(or, equivalently, P = I n ) A = LU We perform LU decomposition is called Gaussian elimination
vA’
Matrix Algebra
In the Ax = b, We must pivot on off-diagonal elements of A to avoid dividing by 0 as well as very small value In case that a 11 ~ a 1n are 0, we can exchange first row and k’th row