Paper 4, Section II, 39D39 \mathrm{D}

Numerical Analysis
Part II, 2012

(i) Formulate the conjugate gradient method for the solution of a system Ax=bA \mathbf{x}=\mathbf{b} with ARn×nA \in \mathbb{R}^{n \times n} and bRn,n>0\mathbf{b} \in \mathbb{R}^{n}, n>0.

(ii) Prove that if the conjugate gradient method is applied in exact arithmetic then, for any x(0)Rn\mathbf{x}^{(0)} \in \mathbb{R}^{n}, termination occurs after at most nn iterations.

(iii) The polynomial p(x)=xm+i=0m1cixip(x)=x^{m}+\sum_{i=0}^{m-1} c_{i} x^{i} is the minimal polynomial of the n×nn \times n matrix AA if it is the polynomial of lowest degree that satisfies p(A)=0.[p(A)=0 . \quad[ Note that mnm \leqslant n.] Give an example of a 3×33 \times 3 symmetric positive definite matrix with a quadratic minimal polynomial.

Prove that (in exact arithmetic) the conjugate gradient method requires at most mm iterations to calculate the exact solution of Ax=bA \mathbf{x}=\mathbf{b}, where mm is the degree of the minimal polynomial of AA.