Page 47 - Matrix Analysis & Applied Linear Algebra
P. 47

1.6 Ill-Conditioned Systems                                                         39

                                           is ill-conditioned by considering the following perturbed system:
                                                                  v − w − x − y − z =0,
                                                                1
                                                              −   v + w − x − y − z =0,
                                                                15
                                                                    1
                                                                  −   v + x − y − z =0,
                                                                    15
                                                                        1
                                                                      −   v + y − z =0,
                                                                       15
                                                                           1
                                                                         −   v + z =1.
                                                                           15
                                    1.6.7. Let f(x) = sin πx on [0, 1]. The object of this problem is to determine
                                           the coefficients α i of the cubic polynomial
                                                                           3
                                                                          !     i
                                                                   p(x)=     α i x
                                                                          i=0
                                           that is as close to f(x) as possible in the sense that
                                                   "  1
                                                                 2
                                               r =   [f(x) − p(x)] dx
                                                    0
                                                                                       #        $ 2
                                                   "  1           3    "  1         "  1  3
                                                                 !         i             !     i
                                                          2
                                                =    [f(x)] dx − 2  α i   x f(x)dx +        α i x  dx
                                                    0                   0            0
                                                                 i=0                     i=0
                                           is as small as possible.
                                              (a) In order to minimize r, impose the condition that ∂r/∂α i =0
                                                  for each i =0, 1, 2, 3, and show this results in a system of linear
                                                  equations whose augmented matrix is [H 4 | b], where H 4 and
                                                  b are given by
                                                               1
                                                                 1   1  1                 2   
                                                                  2   3  4                   π
                                                                                              
                                                              1  1   1  1                 1   
                                                               2  3   4  5                   π
                                                                                              
                                                                               and                .
                                                                                              
                                                       H 4 =                       b = 
                                                              1  1   1  1               1   4 
                                                               3  4   5  6                 π   π 3 
                                                                                          −
                                                                                              
                                                               1  1   1  1                 1   6
                                                               4  5   6  7                 π  −  π 3
                                                  Any matrix H n that has the same form as H 4 is called a
                                                  Hilbert matrix of order n.
                                              (b) Systems involving Hilbert matrices are badly ill-conditioned,
                                                  and the ill-conditioning becomes worse as the size increases. Use
                                                  exact arithmetic with Gaussian elimination to reduce H 4 to tri-
                                                  angular form. Assuming that the case in which n = 4 is typical,
                                                  explain why a general system [H n | b] will be ill-conditioned.
                                                  Notice that even complete pivoting is of no help.
   42   43   44   45   46   47   48   49   50   51   52