Lect 9
Lect 9
Lect 9
Rafikul Alam
Department of Mathematics
IIT Guwahati
For any symmetric matrix A, the product x> Ax is a pure quadratic form f (x1 , . . . , xn ) :
a11 · · · a1n xn n n
> n
. .. .. .. = X X a x x .
x Ax in R x1 · · · xn .. . . . ij i j
an1 ··· ann xn i=1 j=1
Thus, if p is a critical point then f has a local minimum or maximum at p according as the
quadratic form x> Hf (p)x is positive or negative in a neighbourhood of p.
On the other hand, f has a saddle point at p if x> Hf (p)x takes positive and negative values in
a neighbourhood of p.
A real positive definite matrix is also referred to as a symmetric positive definite (SPD) matrix.
A> = A ⇐⇒ A>
m = Am , C = B >, D > = D.
In particular, if A is SPD then ajj = ej> Aej > 0 for j = 1 : n. Also, A is nonsingular (why?).
1 If X ∈ Rn×p with rank(X ) = p then X > AX is SPD. Indeed, for all nonzero y ∈ Rp ,
Xy 6= 0 (why?) and y > (X > AX )y = (Xy )> A(Xy ) > 0 =⇒ X > AX is SPD.
2 Leading principal submatrices of A are SPD, that is, A(1 : j, 1 : j) is SPD for j = 1 : n.
Am B >
3 Let A = . Then S := D − BA−1
m B
>
is the Schur complement of Am . Now
B D
" #" #" #>
B>
Am I 0 Am 0 I 0
=
B D BA−1
m I 0 D − BA−1
m B
>
BA−1
m I
shows that
A is SPD ⇐⇒ Am and S := D − BA−1
m B
>
are SPD.
Theorem: Suppose that all leading principal submatrices A ∈ Rn×n are nonsingular. Then
A = LDV is a unique decomposition of A, where L is unit lower triangular, D is diagonal, and
V is unit upper triangular.
Corollary: If A ∈ Rn×n is symmetric and all leading principal submatrices of A are nonsingular
then A = LDL> is a unique factorization of A, where L is unit lower triangular and D is a
diagonal matrix.
Proof: A = GG > ⇒ x > Ax = x > GG > x = (G > x)> G > x = kG > xk2 > 0 for x 6= 0 ⇒ A is SPD.
Definition: If A is SPD then A = GG > , where G lower triangular with positive diagonals, is
called the Cholesky factorization of A and G is called the Cholesky factor of A.
Example:
>
1 1 1 1 0 0 1 0 0
1 2 2 = 1 1 0 1 1 0 .
1 2 3 1 1 1 1 1 1
a a21 g
Let A := 11 and G := 11 . Then A = GG > yields
a21 a22 g21 g22
2
a11 a21 g g11 g21 g11 g11 g21
= 11 = 2 2 .
a21 a22 g21 g22 g22 g11 g21 g21 + g22
2
Remark: The factorization is possible if a11 > 0 and a22 − g21 > 0.
end
Cost: n3 /3 flops - half the cost of GE.
R. Alam, IIT Guwahati (July-Nov 2023) MA423 MC
Algorithm (inner product)
16 −16 0
Example: Consider −16 41 −5 . Then
0 −5 5
Hence >
16 −16 0 4 4
−16 41 −5 = −4 5 −4 5 .
0 −5 5 0 −1 2 0 −1 2
For k = 1:n
A(k,k) = sqrt(A(k,k));
g = A(k+1:n,k)/A(k,k); A(k+1:n,k) = g;
A(k+1:n, k+1:n) = A(k+1:n, k+1:n)- g*g’;
end
Cost: n3 /3 flops - half the cost of GE.
R. Alam, IIT Guwahati (July-Nov 2023) MA423 MC
Example
25 15 −5 g11 0 0 g11 g21 g31
15 18 0 = g21 g22 0 0 g22 g32
−5 0 11 g31 g32 g33 0 0 g33
5 0 0 5 3 −1
= 3 g22 0 0 g22 g32
−1 g32 g33 0 0 g33
The matlab command chol computes Cholesky factorization of a positive definite matrix A.
More specifically, the commands
and induction on n to prove that Cholesky factorization A = GG > exists and is unique.
***