Académique Documents
Professionnel Documents
Culture Documents
[ http://matrixcookbook.com ]
Kaare Brandt Petersen
Michael Syskind Pedersen
Version: November 14, 2008
What is this? These pages are a collection of facts (identities, approximations, inequalities, relations, ...) about matrices and matters relating to them.
It is collected in this form for the convenience of anyone who wants a quick
desktop reference .
Disclaimer: The identities, approximations and relations presented here were
obviously not invented but collected, borrowed and copied from a large amount
of sources. These sources include similar but shorter notes found on the internet
and appendices in books - see the references for a full list.
Errors: Very likely there are errors, typos, and mistakes for which we apologize and would be grateful to receive corrections at cookbook@2302.dk.
Its ongoing: The project of keeping a large repository of relations involving
matrices is naturally ongoing and the version will be apparent from the date in
the header.
Suggestions: Your suggestion for additional content or elaboration of some
topics is most welcome at cookbook@2302.dk.
Keywords: Matrix algebra, matrix relations, matrix identities, derivative of
determinant, derivative of inverse matrix, differentiate a matrix.
Acknowledgements: We would like to thank the following for contributions
and suggestions: Bill Baxter, Brian Templeton, Christian Rishj, Christian
Schr
oppel Douglas L. Theobald, Esben Hoegh-Rasmussen, Glynne Casteel, Jan
Larsen, Jun Bin Gao, J
urgen Struckmeier, Kamil Dedecius, Korbinian Strimmer, Lars Christiansen, Lars Kai Hansen, Leland Wilkinson, Liguo He, Loic
Thibaut, Miguel Bar
ao, Ole Winther, Pavel Sakov, Stephan Hattinger, Vasile
Sima, Vincent Rabaud, Zhaoshui He. We would also like thank The Oticon
Foundation for funding our PhD studies.
1
CONTENTS
CONTENTS
Contents
1 Basics
1.1 Trace and Determinants . . . . . . . . . . . . . . . . . . . . . . .
1.2 The Special Case 2x2 . . . . . . . . . . . . . . . . . . . . . . . . .
2 Derivatives
2.1 Derivatives
2.2 Derivatives
2.3 Derivatives
2.4 Derivatives
2.5 Derivatives
2.6 Derivatives
2.7 Derivatives
2.8 Derivatives
of
of
of
of
of
of
of
of
a Determinant . . . . . . . . . . . .
an Inverse . . . . . . . . . . . . . . .
Eigenvalues . . . . . . . . . . . . . .
Matrices, Vectors and Scalar Forms
Traces . . . . . . . . . . . . . . . . .
vector norms . . . . . . . . . . . . .
matrix norms . . . . . . . . . . . . .
Structured Matrices . . . . . . . . .
3 Inverses
3.1 Basic . . . . . . . . . . .
3.2 Exact Relations . . . . .
3.3 Implication on Inverses .
3.4 Approximations . . . . .
3.5 Generalized Inverse . . .
3.6 Pseudo Inverse . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5
5
5
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7
7
8
9
9
11
13
13
14
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
16
16
17
19
19
20
20
4 Complex Matrices
23
4.1 Complex Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.2 Higher order and non-linear derivatives . . . . . . . . . . . . . . . 26
4.3 Inverse of complex sum . . . . . . . . . . . . . . . . . . . . . . . 26
5 Solutions and Decompositions
5.1 Solutions to linear equations .
5.2 Eigenvalues and Eigenvectors
5.3 Singular Value Decomposition
5.4 Triangular Decomposition . .
5.5 LU decomposition . . . . . .
5.6 LDM decomposition . . . . .
5.7 LDL decompositions . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
27
27
29
30
32
32
32
32
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 2
CONTENTS
CONTENTS
7 Multivariate Distributions
7.1 Cauchy . . . . . . . . . .
7.2 Dirichlet . . . . . . . . . .
7.3 Normal . . . . . . . . . .
7.4 Normal-Inverse Gamma .
7.5 Gaussian . . . . . . . . . .
7.6 Multinomial . . . . . . . .
7.7 Students t . . . . . . . .
7.8 Wishart . . . . . . . . . .
7.9 Wishart, Inverse . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
36
36
36
36
36
36
36
36
37
38
8 Gaussians
8.1 Basics . . . . . . . .
8.2 Moments . . . . . .
8.3 Miscellaneous . . . .
8.4 Mixture of Gaussians
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
39
39
41
43
44
9 Special Matrices
9.1 Block matrices . . . . . . . . . . . . . . . .
9.2 Discrete Fourier Transform Matrix, The . .
9.3 Hermitian Matrices and skew-Hermitian . .
9.4 Idempotent Matrices . . . . . . . . . . . . .
9.5 Orthogonal matrices . . . . . . . . . . . . .
9.6 Positive Definite and Semi-definite Matrices
9.7 Singleentry Matrix, The . . . . . . . . . . .
9.8 Symmetric, Skew-symmetric/Antisymmetric
9.9 Toeplitz Matrices . . . . . . . . . . . . . . .
9.10 Transition matrices . . . . . . . . . . . . . .
9.11 Units, Permutation and Shift . . . . . . . .
9.12 Vandermonde Matrices . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
45
45
46
47
48
48
50
51
53
54
55
56
57
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
58
58
59
61
61
62
62
63
.
.
.
.
.
.
.
.
.
.
.
.
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
Functions
. . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
A One-dimensional Results
64
A.1 Gaussian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
A.2 One Dimensional Mixture of Gaussians . . . . . . . . . . . . . . . 65
B Proofs and Details
67
B.1 Misc Proofs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 3
CONTENTS
CONTENTS
Matrix
Matrix indexed for some purpose
Matrix indexed for some purpose
Matrix indexed for some purpose
Matrix indexed for some purpose or
The n.th power of a square matrix
The inverse matrix of the matrix A
The pseudo inverse matrix of the matrix A (see Sec. 3.6)
The square root of a matrix (if unique), not elementwise
The (i, j).th entry of the matrix A
The (i, j).th entry of the matrix A
The ij-submatrix, i.e. A with i.th row and j.th column deleted
Vector
Vector indexed for some purpose
The i.th element of the vector a
Scalar
Real part of a scalar
Real part of a vector
Real part of a matrix
Imaginary part of a scalar
Imaginary part of a vector
Imaginary part of a matrix
det(A)
Tr(A)
diag(A)
eig(A)
vec(A)
sup
||A||
AT
AT
A
AH
Determinant of A
Trace of the matrix A
Diagonal matrix of the matrix A, i.e. (diag(A))ij = ij Aij
Eigenvalues of the matrix A
The vector-version of the matrix A (see Sec. 10.2.2)
Supremum of a set
Matrix norm (subscript if any denotes what norm)
Transposed matrix
The inverse of the transposed and vice versa, AT = (A1 )T = (AT )1 .
Complex conjugated matrix
Transposed and complex conjugated matrix (Hermitian)
AB
AB
0
I
Jij
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 4
Basics
(AB)1
(ABC...)1
(AT )1
(A + B)T
(AB)T
(ABC...)T
(AH )1
(A + B)H
(AB)H
(ABC...)H
1.1
B1 A1
...C1 B1 A1
(A1 )T
A T + BT
BT A T
...CT BT AT
(A1 )H
A H + BH
BH A H
...CH BH AH
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
P
Aii
Pi
i = eig(A)
i i ,
Tr(AT )
Tr(BA)
Tr(A) + Tr(B)
Tr(BCA) = Tr(CAB)
Q
i = eig(A)
i i
n
c det(A),
if A Rnn
det(A)
det(A) det(B)
1/ det(A)
det(A)n
1 + uT v
1 + Tr(A),
small
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
(21)
(22)
(23)
(24)
=
=
=
=
=
=
=
=
=
=
1.2
BASICS
=
=
=
=
=
=
=
=
=
=
=
=
=
A=
A11
A21
A12
A22
(25)
(26)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 5
1.2
BASICS
Eigenvalues
2 Tr(A) + det(A) = 0
1 =
Tr(A) +
p
Tr(A)2 4 det(A)
2
1 + 2 = Tr(A)
p
Tr(A)2 4 det(A)
2 =
2
1 2 = det(A)
Tr(A)
Eigenvectors
v1
A12
1 A11
Inverse
A1 =
1
det(A)
v2
A22
A21
A12
2 A11
A12
A11
(27)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 6
DERIVATIVES
Derivatives
x
y
i
x
=
yi
(28)
x
y
=
ij
xi
yj
The following rules are general and very useful when deriving the differential of
an expression ([19]):
A
(X)
(X + Y)
(Tr(X))
(XY)
(X Y)
(X Y)
(X1 )
(det(X))
(ln(det(X)))
XT
XH
2.1
2.1.1
=
=
=
=
=
=
=
=
=
=
=
=
0
(A is a constant)
X
X + Y
Tr(X)
(X)Y + X(Y)
(X) Y + X (Y)
(X) Y + X (Y)
X1 (X)X1
det(X)Tr(X1 X)
Tr(X1 X)
(X)T
(X)H
(29)
(30)
(31)
(32)
(33)
(34)
(35)
(36)
(37)
(38)
(39)
(40)
Derivatives of a Determinant
General form
det(Y)
x
det(Y)
x
x
1 Y
= det(Y)Tr Y
x
" "
#
Y
1 x
= det(Y) Tr Y
x
Y
Y
+Tr Y1
Tr Y1
x
x
#
1 Y
1 Y
Tr Y
Y
x
x
(41)
(42)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 7
2.2
Derivatives of an Inverse
2.1.2
DERIVATIVES
Linear forms
det(X)
X
X det(X)
Xjk
Xik
det(X)(X1 )T
= ij det(X)
(43)
(44)
det(AXB)
X
2.1.3
det(AXB)(X1 )T = det(AXB)(XT )1
(45)
Square forms
(46)
(47)
(48)
2.2
2(X+ )T
= 2XT
=
(X1 )T = (XT )1
= k det(Xk )XT
(49)
(50)
(51)
(52)
Derivatives of an Inverse
(53)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 8
2.3
Derivatives of Eigenvalues
DERIVATIVES
(54)
= XT abT XT
(55)
= det(X1 )(X1 )T
(56)
= (X1 BAX1 )T
(57)
(58)
2.3
Derivatives of Eigenvalues
X
eig(X) =
X
Y
eig(X) =
X
2.4
2.4.1
Tr(X) = I
X
det(X) = det(X)XT
X
(59)
(60)
aT x
x
(61)
= abT
(62)
= baT
(63)
aT XT a
X
aaT
= Jij
(64)
(65)
= im (A)nj
(Jmn A)ij
(66)
= in (A)mj
(Jnm A)ij
(67)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 9
2.4
2.4.2
DERIVATIVES
Second Order
X
Xkl Xmn
Xij
klmn
bT XT Xc
X
(Bx + b)T C(Dx + d)
x
(XT BX)kl
Xij
(XT BX)
Xij
Xkl
(68)
kl
= X(bcT + cbT )
(69)
= BT C(Dx + d) + DT CT (Bx + b)
(70)
(71)
= XT BJij + Jji BX
See Sec 9.7 for useful properties of the Single-entry matrix Jij
xT Bx
x
bT XT DXc
X
(B + BT )x
= DT XbcT + DXcbT
=
(D + DT )(Xb + c)bT
(73)
(74)
(75)
(76)
(77)
(78)
(79)
(80)
= 2b(a xH b)
(81)
(82)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 10
2.5
Derivatives of Traces
DERIVATIVES
n1
Xh
T n T n
a (X ) X b =
X
(83)
r=0
(84)
=
=
2.4.4
s
x
T
Ar +
r
x
T
AT s
xT AT Ax
x xT BT Bx
xT AT AxBT Bx
AT Ax
2
2 T
x BBx
(xT BT Bx)2
(85)
(86)
(87)
Using the above we have for the gradient and the Hessian
f
f
x f =
x
2f
xxT
2.5
= xT Ax + bT x
=
(A + AT )x + b
= A + AT
(88)
(89)
(90)
Derivatives of Traces
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 11
2.5
Derivatives of Traces
2.5.1
DERIVATIVES
First Order
Tr(X) = I
X
Tr(XA) = AT
X
Tr(AXB) = AT BT
X
Tr(AXT B) = BA
X
Tr(XT A) = A
X
Tr(AXT ) = A
X
Tr(A X) = Tr(A)I
X
2.5.2
(91)
(92)
(93)
(94)
(95)
(96)
(97)
Second Order
Tr(X2 )
X
Tr(X2 B)
X
Tr(XT BX)
X
Tr(XBXT )
X
Tr(AXBX)
X
Tr(XT X)
X
Tr(BXXT )
X
Tr(BT XT CXB)
X
Tr XT BXC
X
Tr(AXBXT C)
X
h
i
Tr(X X)
X
2XT
(98)
(XB + BX)T
(99)
= BX + BT X
(100)
= XBT + XB
(101)
= A T X T BT + BT X T A T
(102)
2X
(103)
(B + BT )X
(104)
= CT XBBT + CXBBT
(105)
= BXC + BT XCT
(106)
= AT CT XBT + CAXB
(107)
(108)
Tr(X)Tr(X) = 2Tr(X)I(109)
X
See [7].
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 12
2.6
2.5.3
DERIVATIVES
Higher Order
Tr(Xk )
X
= k(Xk1 )T
k1
X
Tr(AXk ) =
(Xr AXkr1 )T
X
r=0
T T
T
= CXXT CXBBT
X Tr B X CXX CXB
+CT XBBT XT CT X
+CXBBT XT CX
+CT XXT CT XBBT
(110)
(111)
(112)
2.5.4
Other
2.6
2.6.1
2.7
cos(X)T
(117)
||x a||2 =
x
||x a||2
(118)
xa
I
(x a)(x a)T
=
x kx ak2
kx ak2
kx ak32
(119)
||x||22
||xT x||2
=
= 2x
x
x
(120)
2.8
DERIVATIVES
2.7.1
Frobenius norm
||X||2F =
Tr(XXH ) = 2X
(121)
X
X
See (227). Note that this is also a special case of the result in equation 108.
2.8
Assume that the matrix A has some structure, i.e. symmetric, toeplitz, etc.
In that case the derivatives of the previous section does not apply in general.
Instead, consider the following general rule for differentiating a scalar function
f (A)
"
#
T
X f Akl
f
A
df
(122)
=
= Tr
dAij
Akl Aij
A
Aij
kl
(123)
If A has no special structure we have simply Sij = Jij , that is, the structure
matrix is simply the singleentry matrix. Many structures have a representation
in singleentry matrices, see Sec. 9.7.6 for more examples of structure matrices.
2.8.1
(124)
Then the Chain Rule can then be written the following way:
M
g(U)
g(U) X X g(U) ukl
=
=
X
xij
ukl xij
(125)
k=1 l=1
(126)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 14
2.8
2.8.2
DERIVATIVES
Symmetric
(127)
(128)
(129)
2X1 (X1 I)
(130)
Diagonal
= AI
(131)
Toeplitz
Like symmetric matrices and diagonal matrices also Toeplitz matrices has a
special structure which should be taken into account when the derivative with
respect to a matrix with Toeplitz structure.
Tr(AT)
T
Tr(TA)
=
T
(132)
Tr(A)
Tr([AT ]n1 )
Tr([AT ]1n ))
Tr(A)
.
Tr([[AT ]1n ]2,n1 )
.
.
.
A1n
.
.
.
.
.
.
.
.
.
.
Tr([[AT ]1n ]2,n1 )
.
.
An1
.
.
.
.
.
Tr([AT ]n1 )
Tr(A)
Tr([AT ]1n ))
(A)
As it can be seen, the derivative (A) also has a Toeplitz structure. Each value
in the diagonal is the sum of all the diagonal valued in A, the values in the
diagonals next to the main diagonal equal the sum of the diagonal next to the
main diagonal in AT . This result is only valid for the unconstrained Toeplitz
matrix. If the Toeplitz matrix also is symmetric, the same derivative yields
Tr(AT)
Tr(TA)
=
= (A) + (A)T (A) I
T
T
(133)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 15
3
3.1
3.1.1
INVERSES
Inverses
Basic
Definition
(134)
cof(A, 1, 1)
cof(A, 1, n)
.
.
..
.
cof(A) =
cof(A, i, j)
.
cof(A, n, 1)
cof(A, n, n)
(135)
(136)
(137)
Determinant
n
X
j=1
n
X
(138)
(139)
j=1
3.1.4
Construction
1
adj(A)
det(A)
(140)
3.2
Exact Relations
3.1.5
INVERSES
Condition number
The condition number of a matrix c(A) is the ratio between the largest and the
smallest singular value of a matrix (see Section 5.3 on singular values),
c(A) =
d+
d
(141)
The condition number can be used to measure how singular a matrix is. If the
condition number is large, it indicates that the matrix is nearly singular. The
condition number can also be estimated from the matrix norms. Here
c(A) = kAk kA1 k,
(142)
where k k is a norm such as e.g the 1-norm, the 2-norm, the -norm or the
Frobenius norm (see Sec 10.4p
for more on matrix norms).
The 2-norm of A equals (max(eig(AH A))) [12, p.57]. For a symmetric
matrix, this reduces to ||A||2 = max(|eig(A)|) [12, p.394]. If the matrix ia
symmetric and positive definite, ||A||2 = max(eig(A)). The condition number
based on the 2-norm thus reduces to
kAk2 kA1 k2 = max(eig(A)) max(eig(A1 )) =
3.2
3.2.1
max(eig(A))
.
min(eig(A))
Exact Relations
Basic
(AB)1 = B1 A1
3.2.2
(143)
(144)
The Woodbury identity comes in many variants. The latter of the two can be
found in [12]
(A + CBCT )1
(A + UBV)1
(147)
(148)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 17
3.2
Exact Relations
3.2.4
INVERSES
(149)
(150)
(151)
(152)
(153)
(154)
(155)
(156)
1 + dT A+ c
A+ c
(A+ )T d
(I AA+ )c
(I A+ A)T d
(157)
(158)
(159)
(160)
(161)
v
n
w
m
=
=
=
=
=
the solution is given as six different cases, depending on the entities ||w||,
||m||, and . Please note, that for any (column) vector v it holds that v+ =
vT
vT (vT v)1 = ||v||
2 . The solution is:
=
vwT
mnT +
mwT
||w||2
||m||2
||m||2 ||w||2
(162)
(163)
= vv+ A+ (m+ )T nT
1
1
=
vvT A+
mnT
2
||v||
||m||2
(164)
(165)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 18
3.3
Implication on Inverses
INVERSES
||m||2 + T
T +
G = mv A
m+v
(A ) v + n
(166)
Case 4 of 6: If ||w|| 6= 0 and ||m|| = 0 and = 0. Then
G
= A+ nn+ vw+
1
1
=
A+ nnT
vwT
||n||2
||w||2
(167)
(168)
||n||2
G = A+ nwT
(169)
A
w
+
n
n
+
v
3.3
(170)
(171)
Implication on Inverses
(A + B)1 = A1 + B1
If
then
AB1 A = BA1 B
(172)
See [25].
3.3.1
A PosDef identity
(173)
See [30].
3.4
Approximations
(174)
(175)
The following approximation is from [22] and holds when A large and symmetric
A A(I + A)1 A
= I A1
(176)
(177)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 19
3.5
Generalized Inverse
3.5
3.5.1
INVERSES
Generalized Inverse
Definition
A generalized inverse matrix of the matrix A is any matrix A such that (see
[26])
AA A = A
(178)
The matrix A is not unique.
3.6
3.6.1
Pseudo Inverse
Definition
AA+ A = A
A+ AA+ = A+
AA+ symmetric
A+ A symmetric
The matrix A+ is unique and does always exist. Note that in case of complex matrices, the symmetric condition is substituted by a condition of being
Hermitian.
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 20
3.6
Pseudo Inverse
3.6.2
INVERSES
Properties
=
=
=
=
=
6
=
=
=
=
=
=
=
=
=
=
=
=
=
A
(A+ )T
(A+ )H
(A+ )
AH
AT
(1/c)A+
(AT A)+ AT
AT (AAT )+
A+ (AT )+
(AT )+ A+
(AH A)+ AH
AH (AAH )+
A+ (AH )+
(AH )+ A+
(A+ AB)+ (ABB+ )+
A+ [f (AAH ) f (0)I]A
A[f (AH A) f (0)I]A+
(179)
(180)
(181)
(182)
(183)
(184)
(185)
(186)
(187)
(188)
(189)
(190)
(191)
(192)
(193)
(194)
(195)
(196)
where A Cnm .
Assume A to have full rank, then
(AA+ )(AA+ ) =
(A+ A)(A+ A) =
Tr(AA+ ) =
Tr(A+ A) =
AA+
A+ A
rank(AA+ )
rank(A+ A)
(See [26])
(See [26])
(197)
(198)
(199)
(200)
(201)
(202)
Construction
Square
Broad
Tall
rank(A) = n
rank(A) = n
rank(A) = m
A+ = A1
A+ = AT (AAT )1
A+ = (AT A)1 AT
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 21
3.6
Pseudo Inverse
INVERSES
Assume A does not have full rank, i.e. A is nm and rank(A) = r < min(n, m).
The pseudo inverse A+ can be constructed from the singular value decomposition A = UDVT , by
T
A+ = Vr D1
(203)
r Ur
where Ur , Dr , and Vr are the matrices with the degenerated rows and columns
deleted. A different way is this: There do always exist two matrices C n r
and D r m of rank r, such that A = CD. Using these matrices it holds that
A+ = DT (DDT )1 (CT C)1 CT
(204)
See [3].
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 22
COMPLEX MATRICES
Complex Matrices
4.1
Complex Derivatives
(205)
and
df (z)
<(f (z)) =(f (z))
= i
+
dz
=z
=z
or in a more compact form:
f (z)
f (z)
=i
.
=z
<z
(206)
(207)
(208)
(209)
=
=
df (z)
dz
f (z)
f (z)
+i
.
<z
=z
(211)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 23
4.1
Complex Derivatives
COMPLEX MATRICES
=
=
df (Z)
dZ
f (Z)
f (Z)
+i
.
<Z
=Z
(212)
The chain rule is a little more complicated when the function of a complex
u = f (x) is non-analytic. For a non-analytic function, the following chain rule
can be applied ([7])
g(u)
x
=
=
g u
g u
+
u x u x
g u g u
+
u x
u
x
(213)
Notice, if the function is analytic, the second term reduces to zero, and the function is reduced to the normal well-known chain rule. For the matrix derivative
of a scalar function g(U), the chain rule can be written the following way:
T
T
Tr(( g(U)
Tr(( g(U)
g(U)
U ) U)
U ) U )
=
+
.
X
X
X
4.1.2
(214)
If the derivatives involve complex numbers, the conjugate transpose is often involved. The most useful way to show complex derivative is to show the derivative
with respect to the real and the imaginary part separately. An easy example is:
Tr(X )
Tr(XH )
=
<X
<X
Tr(XH )
Tr(X )
=i
i
=X
=X
= I
(215)
= I
(216)
Since the two results have the same sign, the conjugate complex derivative (209)
should be used.
Tr(X)
Tr(XT )
=
<X
<X
Tr(X)
Tr(XT )
i
=i
=X
=X
= I
(217)
= I
(218)
Here, the two results have different signs, and the generalized complex derivative
(208) should be used. Hereby, it can be seen that (92) holds even if X is a
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 24
4.1
Complex Derivatives
COMPLEX MATRICES
complex number.
Tr(AXH )
<X
Tr(AXH )
i
=X
Tr(AX )
<X
Tr(AX )
i
=X
= A
(219)
= A
(220)
= AT
(221)
= AT
(222)
Tr(XXH )
Tr(XH X)
=
<X
<X
H
Tr(XX )
Tr(XH X)
i
=i
=X
=X
2<X
= i2=X
(223)
(224)
By inserting (223) and (224) in (208) and (209), it can be seen that
Tr(XXH )
= X
X
Tr(XXH )
=X
X
(225)
(226)
Since the function Tr(XXH ) is a real function of the complex matrix X, the
complex gradient matrix (212) is given by
Tr(XXH ) = 2
4.1.3
Tr(XXH )
= 2X
X
(227)
=
=
1 det(XH AX)
det(XH AX)
i
2
<X
=X
T
H
H
1 H
det(X AX) (X AX) X A
(228)
=
=
1 det(XH AX)
det(XH AX)
+i
2
<X
=X
det(XH AX)AX(XH AX)1
(229)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 25
4.2
4.2
COMPLEX MATRICES
=
=
4.3
xH AH Ax
x xH BH Bx
AH Ax
xH AH AxBH Bx
2 H
2
x BBx
(xH BH Bx)2
(230)
(231)
Given real matrices A, B find the inverse of the complex sum A + iB. Form
the auxiliary matrices
E = A + tB
F = B tA,
(232)
(233)
=
=
=
=
=
(1 it)(E + iF)1
(234)
1
1
1
1
1
(1 it)((E + FE F) i(E + FE F) FE )(235)
(1 it)(E + FE1 F)1 (I iFE1 )
(236)
1
1
1
1
(E + FE F) ((I tFE ) i(tI + FE ))
(237)
1
1
1
(E + FE F) (I tFE )
i(E + FE1 F)1 (tI + FE1 )
(238)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 26
5
5.1
5.1.1
Assume we have data (xn , yn ) for n = 1, ..., N and are seeking the parameters
a, b R such that yi
= axi + b. With a least squares error function, the optimal
values for a, b can be expressed using the notation
x = (x1 , ..., xN )T
y = (y1 , ..., yN )T
and
Rxx = xT x Rx1 = xT 1
Ryx = yT x Ry1 = yT 1
as
5.1.2
a
b
=
Rxx
Rx1
Rx1
R11
R11 = 1T 1
1
Rx,y
Ry1
(239)
(240)
Solution
Unique solution x
Many solutions x
No solutions x
Standard Square
x = A1 b
(241)
Degenerated Square
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 27
5.1
5.1.5
Cramers rule
The equation
Ax = b,
(242)
where A is square has exactly one solution x if the ith element in x can be
found as
det B
xi =
,
(243)
det A
where B equals A, but the ith column in A has been substituted by b.
5.1.6
Over-determined Rectangular
x = (AT A)1 AT b = A+ b
(244)
xmin = A+ b
(245)
Now xmin is the vector x which minimizes ||Ax b||2 , i.e. the vector which is
least wrong. The matrix A+ is the pseudo-inverse of A. See [3].
5.1.7
Under-determined Rectangular
xmin = AT (AAT )1 b
(246)
The equation have many solutions x. But xmin is the solution which minimizes
||Ax b||2 and also the solution with the smallest norm ||x||2 . The same holds
for a matrix version: Assume A is n m, X is m n and B is n n, then
AX = B
Xmin = A+ B
(247)
The equation have many solutions X. But Xmin is the solution which minimizes
||AX B||2 and also the solution with the smallest norm ||X||2 . See [3].
Similar but different: Assume A is square n n and the matrices B0 , B1
are n N , where N > n, then if B0 has maximal rank
AB0 = B1
(248)
where Amin denotes the matrix which is optimal in a least square sense. An
interpretation is that A is the linear approximation which maps the columns
vectors of B0 into the columns vectors of B1 .
5.1.8
A=0
(249)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 28
5.2
5.1.9
If A is symmetric, then
xT Ax = 0,
5.1.10
A=0
(250)
(251)
(252)
Sec 10.2.1 and 10.2.2 for details on the Kronecker product and the vec operator.
5.1.11
Encapsulating Sum
P
n An XBn
= C
1
P T
vec(X) =
vec(C)
n Bn A n
(253)
(254)
See Sec 10.2.1 and 10.2.2 for details on the Kronecker product and the vec
operator.
5.2
5.2.1
(D)ij = ij i ,
(255)
(256)
General Properties
(257)
(258)
(259)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 29
5.3
5.2.3
Symmetric
=
=
=
=
I
R
P
(i.e. V is orthogonal)
(i.e. i is real)
p
i i
1 + ci
i c
1
i
(260)
(261)
(262)
(263)
(264)
(265)
(266)
Characteristic polynomial
= det(A I)
= n g1 n1 + g2 n2 ... + (1)n gn
(267)
(268)
Note that the coefficients gj for j = 1, ..., n are the n invariants under rotation
of A. Thus, gj is the sum of the determinants of all the sub-matrices of A taken
j rows and columns at a time. That is, g1 is the trace of A, and g2 is the sum
of the determinants of the n(n 1)/2 sub-matrices that can be formed from A
by deleting all but two rows and columns, and so on see [17].
5.3
5.3.1
U = p
eigenvectors of AAT
D =
diag(eig(AAT ))
V = eigenvectors of AT A
(269)
nn
nm
mm
(270)
(271)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 30
5.3
5.3.2
UT
(272)
where D is diagonal with the square root of the eigenvalues of AAT , V is the
eigenvectors of AAT and UT is the eigenvectors of AT A.
5.3.3
(273)
Rectangular decomposition I
Assume A is n m, V is n n, D is n n, UT is n m
A
D
UT
= V
,
(274)
where D is diagonal with the square root of the eigenvalues of AAT , V is the
eigenvectors of AAT and UT is the eigenvectors of AT A.
5.3.5
Rectangular decomposition II
Assume A is n m, V is n m, D is m m, UT is m m
D UT
A
V
=
5.3.6
(275)
Assume A is n m, V is n n, D is n m, UT is m m
UT
,
A
D
= V
(276)
where D is diagonal with the square root of the eigenvalues of AAT , V is the
eigenvectors of AAT and UT is the eigenvectors of AT A.
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 31
5.4
Triangular Decomposition
5.4
Triangular Decomposition
5.5
LU decomposition
(277)
Cholesky-decomposition
(278)
5.6
LDM decomposition
(279)
where L, M are unique unit lower triangular matrices and D is a unique diagonal
matrix.
5.7
LDL decompositions
The LDL decomposition are special cases of the LDM decomposition. Assume
A is a non-singular symmetric definite square matrix, then
A = LDLT = LT DL
(280)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 32
6.1
Definition of Moments
Mean
(281)
Covariance
(282)
(283)
or alternatively as
6.1.3
Third moments
(284)
h
i
(3) (3)
M3 = m::1 m::2 ...m(3)
::n
(285)
as
where : denotes all elements within the given index. M3 can alternatively be
expressed as
M3 = h(x m)(x m)T (x m)T i
(286)
6.1.4
Fourth moments
mijkl = h(xi hxi i)(xj hxj i)(xk hxk i)(xl hxl i)i
(287)
as
i
h
(4)
(4)
(4)
(4)
(4)
(4)
(4)
(4)
M4 = m::11 m::21 ...m::n1 |m::12 m::22 ...m::n2 |...|m::1n m::2n ...m(4)
::nn
(288)
or alternatively as
M4 = h(x m)(x m)T (x m)T (x m)T i
(289)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 33
6.2
6.2
6.2.1
Linear Forms
(290)
(291)
(292)
(293)
(294)
(295)
Quadratic Forms
=
=
=
=
=
=
=
=
=
=
=
(298)
(299)
(300)
(301)
(302)
(303)
(304)
(305)
(306)
(307)
(308)
See [7].
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 34
6.3
6.2.3
Cubic Forms
= Adiag(BT C)v3
+Tr(BMCT )(Am + a)
+AMCT (Bm + b)
+(AMBT + (Am + a)(Bm + b)T )(Cm + c)
E[xxT x] = v3 + 2Mm + (Tr(M) + mT m)m
E[(Ax + a)(Ax + a)T (Ax + a)] = Adiag(AT A)v3
+[2AMAT + (Ax + a)(Ax + a)T ](Am + a)
+Tr(AMAT )(Am + a)
E[(Ax + a)bT (Cx + c)(Dx + d)T ] = (Ax + a)bT (CMDT + (Cm + c)(Dm + d)T )
+(AMCT + (Am + a)(Cm + c)T )b(Dm + d)T
+bT (Cm + c)(AMDT (Am + a)(Dm + d)T )
6.3
=
=
=
=
wT m
wT M2 w
wT M3 w w
wT M4 w w w
(309)
(310)
(311)
(312)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 35
MULTIVARIATE DISTRIBUTIONS
Multivariate Distributions
7.1
Cauchy
( 1+P
det()1/2
2 )
(1/2) 1 + (t )T 1 (t )(1+P )/2
(313)
where is the location, is positive definite, and denotes the gamma function. The Cauchy distribution is a special case of the Student-t distribution.
7.2
Dirichlet
p Y
p
p 1
x
p(x|) = QP
p
(
)
p
p
p
7.3
Normal
7.4
Normal-Inverse Gamma
7.5
Gaussian
See sec. 8.
7.6
Multinomial
If the vector n contains counts, i.e. (n)i 0, 1, 2, ..., then the discrete multinomial disitrbution for n is given by
d
P (n|a, n) =
7.7
d
X
Y
n!
ani ,
n1 ! . . . n d ! i i
ni = n
(314)
ai = 1.
Students t
( +P
det()1/2
2 )
(/2) 1 + 1 (t )T 1 (t )(+P )/2
(315)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 36
7.8
Wishart
MULTIVARIATE DISTRIBUTIONS
Mean
E(t) = ,
7.7.2
(316)
Variance
cov(t) =
7.7.3
>1
,
2
>2
(317)
Mode
The notion mode meaning the position of the most probable value
mode(t) =
7.7.4
(318)
= N P/2
P
Y
[( + P p + 1)/2]
[( p + 1)/2]
p=1
det()/2 det()N/2
(+P )/2
det 1 + (T M)1 (T M)T
(319)
where M is the location, is the rescaling matrix, is positive definite, is
the degrees of freedom, and denotes the gamma function.
7.8
Wishart
2mP/2 P (P 1)/4
1
QP
p
[ 12 (m + 1 p)]
(320)
Mean
E(M) = m
(321)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 37
7.9
Wishart, Inverse
7.9
MULTIVARIATE DISTRIBUTIONS
Wishart, Inverse
The (normal) Inverse Wishart distribution for M RP P , M is positive definite, where m can be regarded as a degree of freedom parameter [11]
p(M|, m)
2mP/2 P (P 1)/4
1
QP
p
[ 12 (m + 1 p)]
(322)
Mean
E(M) =
1
mP 1
(323)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 38
8
8.1
8.1.1
GAUSSIANS
Gaussians
Basics
Density and normalization
1
exp (x m)T 1 (x m)
2
det(2)
1
(324)
p
det(2)
1 T 1
=
det(2) exp m m
2
p
1 T T
1
=
det(2A ) exp c A c
2
p
= p(x)1 (x m)
= p(x) 1 (x m)(x m)T 1 1
(325)
(326)
Marginal Distribution
Assume x Nx (, ) where
xa
a
x=
=
xb
b
=
a
Tc
c
b
(327)
then
p(xa )
p(xb )
= Nxa (a , a )
= Nxb (b , b )
(328)
(329)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 39
8.1
Basics
8.1.3
GAUSSIANS
Conditional Distribution
Assume x Nx (, ) where
xa
a
x=
=
xb
b
=
a
Tc
c
b
(330)
then
a)
p(xa |xb ) = Nxa (
a ,
b)
p(xb |xa ) = Nxb (
b ,
n
a = a + c 1
b (xb b ) (331)
a = a c 1 T
c
b
n
T
1
b = b + c a (xa a )
(332)
b = b T 1 c
c
a
Note, that the covariance matrices are the Schur complement of the block matrix, see 9.1.5 for details.
8.1.4
Linear combination
(333)
8.1.5
8.1.6
(334)
If A is symmetric, then
1
1
1
xT Ax + bT x = (x A1 b)T A(x A1 b) + bT A1 b
2
2
2
1
1
1
T
T
1
T
1
Tr(X AX) + Tr(B X) = Tr[(X A B) A(X A B)] + Tr(BT A1 B)
2
2
2
8.1.7
(335)
(336)
(337)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 40
8.2
Moments
1
c
mc
C
GAUSSIANS
1
= 1
(338)
1 + 2
1
1 1
1
1
= (1 + 2 ) (1 m1 + 2 m2 )
(339)
1 T 1
1
1 1
1
(m + mT2 1
(1
=
2 )(1 + 2 )
1 m1 + 2 m2 )(340)
2 1 1
1
T 1
m
+
m
m
(341)
mT1 1
1
2
2
1
2
2
8.1.8
(342)
(343)
(344)
1
= 1
(345)
1 + 2
1
1 1
1
1
= (1 + 2 ) (1 M1 + 2 M2 )
(346)
i
1 h 1
1
1 1
1
T
=
Tr (1 M1 + 1
(1
2 M2 ) (1 + 2 )
1 M1 + 2 M2 )
2
1
T 1
Tr(MT1 1
(347)
1 M1 + M2 2 M2 )
2
mc
c
(348)
= Nm1 (m2 , (1 + 2 ))
1
1
= p
exp (m1 m2 )T (1 + 2 )1 (m1 m2 )
2
det(2(1 + 2 ))
1 1
1
= (1
(1
1 + 2 )
1 m1 + 2 m2 )
1 1
= (1
1 + 2 )
8.2
8.2.1
Moments
Mean and covariance of linear forms
(349)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 41
8.2
Moments
GAUSSIANS
(350)
(351)
(352)
(353)
2 4 Tr(A2 ) + 4 2 mT A2 m
(358)
Cubic forms
E[xbT xxT ]
8.2.4
(359)
(360)
= 2( + mmT )2 + mT m( mmT )
+Tr()( + mmT )
E[xxT AxxT ] = ( + mmT )(A + AT )( + mmT )
+mT Am( mmT ) + Tr[A]( + mmT )
T
T
E[x xx x] = 2Tr(2 ) + 4mT m + (Tr() + mT m)2
E[xT AxxT Bx] = Tr[A(B + BT )] + mT (A + AT )(B + BT )m
+(Tr(A) + mT Am)(Tr(B) + mT Bm)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 42
8.3
Miscellaneous
GAUSSIANS
Moments
E[x]
k m k
(361)
Cov(x)
XX
8.3
8.3.1
k k0 (k + mk mTk mk mTk0 )
(362)
k0
Miscellaneous
Whitening
(363)
1/2
Note that
means the matrix which fulfils
and is unique since is positive definite.
8.3.2
1/2
(364)
= , and that it exists
(365)
8.4
Mixture of Gaussians
8.3.3
GAUSSIANS
Entropy
8.4
8.4.1
(366)
Mixture of Gaussians
Density
K
X
k=1
1
exp (x mk )T 1
(x
m
)
k
k
2
det(2k )
k p
(367)
Derivatives
P
Defining p(s) = k k Ns (k , k ) one get
ln p(s)
j
=
=
ln p(s)
j
=
=
ln p(s)
j
=
=
j Ns (j , j )
P
ln[j Ns (j , j )]
(368)
N
(
,
)
j
k
s
k
k
k
j Ns (j , j ) 1
P
(369)
k k Ns (k , k ) j
j Ns (j , j )
P
ln[j Ns (j , j )]
(370)
k k Ns (k , k ) j
j Ns (j , j ) 1
P
j (s j )
(371)
k k Ns (k , k )
j Ns (j , j )
P
ln[j Ns (j , j )]
(372)
N
(
,
k
j
k
k k s
j Ns (j , j ) 1
1
T 1
P
1
j + j (s j )(s j ) j (373)
2
N
(
,
)
k
k
k k s
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 44
9
9.1
SPECIAL MATRICES
Special Matrices
Block matrices
Multiplication
The Determinant
C1
C2
as
det
9.1.3
A12
A22
A11
A21
(374)
(375)
The Inverse
C1
C2
as
=
9.1.4
A12
A22
A11
A21
1
=
C1
1
1
C2 A21 A1
11
1
1
1
A1
11 + A11 A12 C2 A21 A11
1
1
A22 A21 C1
(376)
(377)
1
A1
11 A12 C2
1
C2
1
C1
1 A12 A22
1
1
1
A22 + A22 A21 C1 A12 A1
22
Block diagonal
A11
0
A11
0
0
A22
0
A22
1
=
(A11 )1
0
0
(A22 )1
(378)
=
det(A11 ) det(A22 )
(379)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 45
9.2
9.1.5
SPECIAL MATRICES
Schur complement
A12
A22
is the matrix
A11 A12 A1
22 A21
that is, what is denoted C2 above. Using the Schur complement, one can rewrite
the inverse of a block matrix
1
A11 A12
A21 A22
1
I
0
0
(A11 A12 A1
I A12 A1
22 A21 )
22
=
A1
I
0
I
0
A1
22 A21
22
The Schur complement is useful when solving linear systems of the form
A11 A12
x1
b1
=
A21 A22
x2
b2
which has the following equation for x1
1
(A11 A12 A1
22 A21 )x1 = b1 A12 A22 b2
When the appropriate inverses exists, this can be solved for x1 which can then
be inserted in the equation for x2 to solve for x2 .
9.2
N
1
X
x(n)WNkn .
(381)
n=0
N 1
1 X
X(k)WNkn .
N
(382)
k=0
The DFT of the vector x = [x(0), x(1), , x(N 1)]T can be written in matrix
form as
X = WN x,
(383)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 46
9.3
SPECIAL MATRICES
(384)
j2
N
1
W
N N
= NI
= WH
N
(385)
(386)
(387)
, then [23]
m+N/2
WN
= WNm
(388)
(389)
9.3
x Cn1
(390)
(391)
Note that
A = B + iC
where B, C are hermitian, then
B=
9.3.1
A + AH
,
2
C=
A AH
2i
Skew-Hermitian
(392)
(393)
(394)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 47
9.4
Idempotent Matrices
9.4
SPECIAL MATRICES
Idempotent Matrices
A matrix A is idempotent if
AA = A
Idempotent matrices A and B, have the following properties
An = A,
forn = 1, 2, 3, ...
IA
is idempotent
AH
is idempotent
H
IA
is idempotent
If AB = BA AB is idempotent
rank(A) = Tr(A)
A(I A) = 0
(I A)A = 0
A+ = A
f (sI + tA) = (I A)f (s) + Af (s + t)
(395)
(396)
(397)
(398)
(399)
(400)
(401)
(402)
(403)
(404)
Nilpotent
A matrix A is nilpotent if
A2 = 0
A nilpotent matrix has the following property:
f (sI + tA)
9.4.2
(405)
Unipotent
A matrix A is unipotent if
AA = I
A unipotent matrix has the following property:
f (sI + tA)
9.5
(406)
Orthogonal matrices
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 48
9.5
Orthogonal matrices
SPECIAL MATRICES
=
=
=
=
=
QT
Q
I
I
1
Ortho-Sym
= I
=
QT+
(407)
(408)
=
=
9.5.2
1 + (1)k
1 + (1)k+1
I+
Q+
2
2
1 + cos(k)
1 cos(k)
I+
Q+
2
2
(409)
(410)
Ortho-Skew
= I
=
QH
(411)
(412)
=
=
9.5.3
ik + (i)k
ik (i)k
Ii
Q
2
2
cos(k )I + sin(k )Q
2
2
(413)
(414)
Decomposition
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 49
9.6
9.6
9.6.1
SPECIAL MATRICES
x 6= 0
(416)
(417)
Eigenvalues
eig( A+A
)>0
2
A+AH
eig( 2 ) 0
(418)
Trace
Tr(A) > 0
Tr(A) 0
(419)
Inverse
Diagonal
Decomposition I
Decomposition II
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 50
9.7
9.6.8
SPECIAL MATRICES
AX = 0
Rank of product
Outer Product
Small pertubations
Hadamard inequality
Assume that P = AAT and Q = BBT are semi positive definite matrices, it
then holds that
P Q = RRT
where the columns of R are constructed as follows: ri+(j1)NA = ai bj , for
i = 1, 2, ..., NA and j = 1, 2, ..., NB . The result is unpublished, but reported by
Pavel Sakov and Craig Bishop.
9.7
9.7.1
The single-entry matrix Jij Rnn is defined as the matrix which is zero
everywhere except in the entry (i, j) in which it is 1. In a 4 4 example one
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 51
9.7
SPECIAL MATRICES
might have
J23
0
0
=
0
0
0
0
0
0
0
1
0
0
0
0
0
0
(420)
The single-entry matrix is very useful when working with derivatives of expressions involving matrices.
9.7.2
... 0
(421)
i.e. an n p matrix of zeros with the i.th column of A in place of the j.th
column. Assume A to be n m and Jij to be p n
0
..
.
(422)
Jij A =
Aj
0
.
..
0
i.e. an p m matrix of zeros with the j.th row of A in the placed of the i.th
row.
9.7.3
(AJij B)kl
(423)
(424)
Aik Bjl =
Aki Blj =
9.7.4
(A ei eTj B)kl
(Aei eTj BT )kl
=
=
ij
(A J B)kl
ij
(AJ B )kl
(425)
(426)
If i = j
Jij Jij = Jij
Jij (Jij )T = Jij
If i 6= j
Jij Jij = 0
Jij (Jij )T = Jii
(Jij )T (Jij )T = 0
(Jij )T Jij = Jjj
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 52
9.8
Symmetric, Skew-symmetric/Antisymmetric
9.7.5
SPECIAL MATRICES
(427)
(428)
(429)
(430)
9.7.6
(431)
(432)
Structure Matrices
(433)
Sij = Jij
(434)
(435)
If A is symmetric then
9.8
9.8.1
Symmetric, Skew-symmetric/Antisymmetric
Symmetric
(436)
Symmetric matrices have many important properties, e.g. that their eigenvalues
are real and eigenvectors orthogonal.
9.8.2
Skew-symmetric/Antisymmetric
The antisymmetric matrix is also known as the skew symmetric matrix. It has
the following property from which it is defined
A = AT
(437)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 53
9.9
Toeplitz Matrices
SPECIAL MATRICES
Hereby, it can be seen that the antisymmetric matrices always have a zero
diagonal. The n n antisymmetric matrices also have the following properties.
det(AT )
det(A)
=
=
(438)
(439)
Decomposition
9.9
A AT
A + AT
+
= A+ + A
2
2
(441)
Toeplitz Matrices
t0
t1 tn1
t11 t12 t1n
..
..
..
..
t1
t21 . . . . . .
.
.
.
.
(442)
T= .
=
.
.
.
.
.
..
..
..
.. t
..
..
t1
12
t(n1) t1
t0
tn1 t21 t11
A Toeplitz matrix is persymmetric. If a matrix is persymmetric (or orthosymmetric), it means that the matrix is symmetric about its northeast-southwest
diagonal (anti-diagonal) [12]. Persymmetric matrices is a larger class of matrices, since a persymmetric matrix not necessarily has a Toeplitz structure. There
are some special cases of Toeplitz matrices. The symmetric Toeplitz matrix is
given by:
t0
t1 tn1
..
.. ..
.
.
t1
.
T=
(443)
..
.
.
.
.
.
.
.
t1
t(n1) t1
t0
The circular Toeplitz matrix:
t0
TC =
tn
..
.
t1
t1
..
.
..
..
.
..
.
.
tn1
tn1
..
.
t1
t0
(444)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 54
9.10
Transition matrices
t0 t1 tn1
..
0 ... ...
.
TU = .
.
.
.. ..
..
t1
0 0
t0
and the lower triangular Toeplitz matrix:
t0
0
.
.. ...
t1
TL =
..
..
..
.
.
.
t(n1) t1
9.9.1
SPECIAL MATRICES
(445)
0
..
.
0
t0
(446)
The Toeplitz matrix has some computational advantages. The addition of two
Toeplitz matrices can be done with O(n) flops, multiplication of two Toeplitz
matrices can be done in O(n ln n) flops. Toeplitz equation systems can be solved
in O(n2 ) flops. The inverse of a positive definite Toeplitz matrix can be found
in O(n2 ) flops too. The inverse of a Toeplitz matrix is persymmetric. The
product of two lower triangular Toeplitz matrices is a Toeplitz matrix. More
information on Toeplitz matrices and circulant matrices can be found in [13, 7].
9.10
Transition matrices
The transition matrix usually describes the probability of moving from state i
to j in one step and is closely related to markov processes. Transition matrices
have the following properties
Prob[i j in 1 step] = (P)ij
Prob[i j in 2 steps] = (P2 )ij
Prob[i j in k steps] = (Pk )ij
If all rows are identical Pn = P
P = ,
is called invariant
where is a so-called stationary probability vector, i.e., 0 i 1 and
1.
(447)
(448)
(449)
(450)
(451)
P
i =
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 55
9.11
9.11
SPECIAL MATRICES
9.11.1
Unit vector
Let ei Rn1 be the ith unit vector, i.e. the vector which is zero in all entries
except the ith at which it is 1.
9.11.2
i.th row of A
j.th column of A
9.11.3
(452)
(453)
Permutations
0 1
P= 1 0
0 0
matrix, e.g.
0
0 = e2
1
eT2
= eT1
eT3
e1
e3
(454)
(455)
eT2 A
PA = eT1 A
eT3 A
AP =
Ae2
Ae1
Ae3
(456)
That is, the first is a matrix which has columns of A but in permuted sequence
and the second is a matrix which has the rows of A but in the permuted sequence.
9.11.4
0
1
L=
0
0
0
0
0
1
0
0
0
0
(457)
i.e. a matrix of zeros with one on the sub-diagonal, (L)ij = i,j+1 . With some
signal xt for t = 1, ..., N , the n.th power of the lag operator shifts the indices,
i.e.
n 0
for t = 1, .., n
(Ln x)t =
(458)
xtn for t = n + 1, ..., N
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 56
9.12
Vandermonde Matrices
SPECIAL MATRICES
A related but slightly different matrix is the recurrent shifted operator defined
on a 4x4 example by
0 0 0 1
= 1 0 0 0
L
(459)
0 1 0 0
0 0 1 0
ij = i,j+1 + i,1 j,dim(L) . On a signal x it has the
i.e. a matrix defined by (L)
effect
n x)t = xt0 , t0 = [(t n) mod N ] + 1
(L
(460)
is like the shift operator L except that it wraps the signal as if it
That is, L
was periodic and shifted (substituting the zeros with the rear end of the signal).
is invertible and orthogonal, i.e.
Note that L
1 = L
T
L
9.12
(461)
Vandermonde Matrices
1 v1 v12 v1n1
1 v2 v22 v n1
2
V= . .
..
..
.
.
. .
.
.
1 vn vn2 vnn1
(462)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 57
10
10
10.1
10.1.1
10.1.2
(464)
Consider some scalar function f (x) which takes the vector x as an argument.
This we can Taylor expand around x0
1
f (x)
= f (x0 ) + g(x0 )T (x x0 ) + (x x0 )T H(x0 )(x x0 )
2
where
g(x0 ) =
10.1.3
f (x)
x x0
H(x0 ) =
(465)
2 f (x)
xxT x0
As for analytical functions in one dimension, one can define a matrix function
for square matrices X by an infinite series
f (X) =
cn Xn
(466)
n=0
P
assuming the limit exists and is finite. If the coefficients cn fulfils n cn xn < ,
then one can prove that the above series exists and is finite, see [1]. Thus for
any analytical function f (x) there exists a corresponding matrix function f (x)
constructed by the Taylor expansion. Using this one can prove the following
results:
1) A matrix A is a zero of its own characteristic polynomium [1]:
X
p() = det(I A) =
cn n
p(A) = 0
(467)
n
f (A) = Uf (B)U1
(468)
if
|A| < 1
(469)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 58
10.2
10.1.4
10
In analogy to the ordinary scalar exponential function, one can define exponential and logarithmic matrix functions:
eA
X
1
1 n
A = I + A + A2 + ...
n!
2
n=0
(470)
eA
X
1
1
(1)n An = I A + A2 ...
n!
2
n=0
(471)
etA
X
1
1
(tA)n = I + tA + t2 A2 + ...
n!
2
n=0
(472)
X
(1)n1 n
1
1
A = A A2 + A3 ...
n
2
3
n=1
(473)
ln(I + A)
= eA+B
= eA
if
AB = BA
= AetA = etA A,
=
tR
Tr(AetA )
= eTr(A)
(478)
Trigonometric Functions
cos(A)
10.2.1
(476)
(477)
X
(1)n A2n+1
1
1
= A A3 + A5 ...
sin(A)
(2n
+
1)!
3!
5!
n=0
10.2
(474)
(475)
X
(1)n A2n
1
1
= I A2 + A4 ...
(2n)!
2!
4!
n=0
(479)
(480)
AB=
(481)
..
..
.
.
Am1 B Am2 B
... Amn B
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 59
10.2
10
AB+AC
BA
in general
(A B) C
A B (A B)
A T BT
AC BD
A1 B1
A + B+
rank(A)rank(B)
Tr(A)Tr(B) = Tr(A B )
= det(A)rank(B) det(B)rank(A)
= {eig(B A)}
if A, B are square
T
= {eig(A)eig(B) }
if A, B are symmetric and square
= eig(A) eig(B)
=
6=
=
=
=
=
=
=
=
=
(482)
(483)
(484)
(485)
(486)
(487)
(488)
(489)
(490)
(491)
(492)
(493)
(494)
(495)
Where {i } denotes the set of values i , that is, the values in no particular
order or structure, and A denotes the diagonal matrix with the eigenvalues of
A.
10.2.2
The vec-operator applied on a matrix A stacks the columns into a vector, i.e.
for a 2 2 matrix
A11
A21
A11 A12
A=
vec(A) =
A12
A21 A22
A22
Properties of the vec-operator include (see [19])
vec(AXB) = (BT A)vec(X)
Tr(AT B) = vec(A)T vec(B)
vec(A + B) = vec(A) + vec(B)
vec(A) = vec(A)
(496)
(497)
(498)
(499)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 60
10.3
Vector Norms
10.3
10.3.1
10
Vector Norms
Examples
||x||1
||x||22
||x||p
|xi |
(500)
= x x
"
#1/p
X
p
=
|xi |
(501)
i
H
(502)
||x||
max |xi |
(503)
10.4
10.4.1
Matrix Norms
Definitions
(504)
(505)
(506)
(507)
| ||x|| = 1}
(508)
where || || ont the left side is the induced matrix norm, while || || on the right
side denotes the vector norm. For induced norms it holds that
||I|| = 1
||Ax|| ||A|| ||x||,
||AB|| ||A|| ||B||,
10.4.3
for all A, x
for all A, B
(509)
(510)
(511)
Examples
||A||1
||A||2
||A||p
||A||
X
max
|Aij |
j
i
q
=
max eig(AH A)
= ( max ||Ax||p )1/p
||x||p =1
X
= max
|Aij |
=
(512)
(513)
(514)
(515)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 61
10.5
Rank
10
||A||F
sX
|Aij |2 =
q
Tr(AAH )
(Frobenius)
(516)
ij
||A||max
max |Aij |
||A||KF
= ||sing(A)||1
(517)
ij
(Ky Fan)
(518)
Inequalities
m
n
mn
mn
mnd
||A||1
1
||A||
1
m
n
n
n
nd
m
md
||A||2
1
m
n
d
d
||A||F
1
m
n
1
||A||KF
1
m
n
1
1
m ||A||
(519)
10.4.5
Condition Number
p
The 2-norm of A equals (max(eig(AT A))) [12, p.57]. For a symmetric, positive definite matrix, this reduces to max(eig(A)) The condition number based
on the 2-norm thus reduces to
kAk2 kA1 k2 = max(eig(A)) max(eig(A1 )) =
10.5
10.5.1
max(eig(A))
.
min(eig(A))
(520)
Rank
Sylvesters Inequality
If A is m n and B is n r, then
rank(A) + rank(B) n rank(AB) min{rank(A), rank(B)}
10.6
(521)
1
p(A1 x)
det(A)
(522)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 62
10.7
Miscellaneous
10
(523)
See [9].
10.7
Miscellaneous
(524)
It holds that
A is positive definite
(525)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 63
A
A.1
A.1.1
ONE-DIMENSIONAL RESULTS
One-dimensional Results
Gaussian
Density
(x )2
p(x) =
exp
2 2
2 2
A.1.2
A.1.3
Normalization
Z
(s)2
e 22 ds =
2 2
r
2
Z
b 4ac
(ax2 +bx+c)
e
dx =
exp
a
4a
2
r
Z
c1 4c2 c0
c2 x2 +c1 x+c0
e
dx =
exp
c2
4c2
(526)
(527)
(528)
(529)
Derivatives
p(x)
ln p(x)
p(x)
ln p(x)
A.1.4
(x )
2
(x )
=
2
1 (x )2
= p(x)
2
1 (x )2
=
2
= p(x)
(530)
(531)
(532)
(533)
or
b=
1 c1
2 c2
w=
1 c21
+ c0
4 c2
1
(x )2 + d
2 2
1
c2
2 =
d = c0 1
2c2
4c2
c2 x2 + c1 x + c0 =
=
A.1.5
c1
2c2
Moments
or p(x) = C exp(c2 x2 + c1 x)
(534)
A.2
hxi
hx2 i
2 + 2
hx3 i
3 2 + 3
hx4 i
4 + 62 2 + 3 4
c1
2c2
2
c1
1
2c2 + h 2c2
i
c21
c1
3
2
(2c )
2c2
2 4
2
c1
c1
+
6
2c2
2c2
1
2c2
+3
1
2c2
2
h(x ) i
h(x )3 i
4
h(x ) i
=
=
=
=
=
2
0h
=
=
1
2c2
1
2c2
i2
c1
2
n
n
exp(c2 x + c1 x)x dx = Zhx i =
exp
hxn i
(535)
c2
4c2
From the un-centralized moments one can derive other entities like
A.2
A.2.1
hx2 i hxi2
hx3 i hx2 ihxi
=
=
2
2 2
=
=
hx4 i hx2 i2
2 4 + 42 2
c2
1 4 2c12
K
X
k
A.2.2
1
2c2
2c1
(2c2 )2
2
(2c2 )2
1 (s k )2
p k exp
2
k2
2k2
(536)
Moments
k hxn ik
(537)
where hik denotes average with respect to the k.th component. We can calculate
the first four moments from the densities
X
1
1 (x k )2
p(x) =
k p
exp
(538)
2
k2
2k2
k
X
p(x) =
k Ck exp ck2 x2 + ck1 x
(539)
k
as
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 65
A.2
hx3 i
hx4 i
hxi
2
hx i
k k
2
k k (k
2k )
2
3
k k (3k k + k )
P
4
2 2
4
k k (k + 6k k + 3k )
ck1
2ck2
1
2ck2
i
+
ck1
2ck2
2
h
ii
c2k1
ck1
3
2
k
k
(2ck2 )
2ck2
2
2
P
c2k1
ck1
1
6 2ck2 + 3
k k
2ck2
2ck2
h
hx i
hx3 i
=
=
hx i
0
P
2
k k k
0
P
4
k k 3k
=
=
=
=
0
P
k k
0
P
k k 3
1
2ck2
1
2ck2
i2
From the un-centralized moments one can derive other entities like
2
P
2
hx2 i hxi2
=
0 k k 0 k + k k k 0
k,k
2
P
3
2
2
hx3 i hx2 ihxi =
k,k0 k k0 3k k + k (k + k )k0
P
4
2 2
4
2
2
2
2
hx4 i hx2 i2
=
k,k0 k k0 k + 6k k + 3k (k + k )(k0 + k0 )
A.2.3
Derivatives
P
Defining p(s) = k k Ns (k , k2 ) we get for a parameter j of the j.th component
j Ns (j , j2 ) ln(j Ns (j , j2 ))
ln p(s)
=P
(540)
2
j
j
k k Ns (k , k )
that is,
ln p(s)
j
ln p(s)
j
ln p(s)
j
j Ns (j , j2 ) 1
P
2
k k Ns (k , k ) j
(541)
j Ns (j , j2 ) (s j )
P
2
j2
k k Ns (k , k )
"
#
j Ns (j , j2 ) 1 (s j )2
P
1
2
j2
k k Ns (k , k ) j
(542)
(543)
Note thatP
k must be constrained to be proper ratios. Defining the ratios by
j = erj / k erk , we obtain
ln p(s) X ln p(s) l
=
rj
l rj
l
where
l
= l (lj j )
rj
(544)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 66
B
B.1
B.1.1
Xij
u1 ,...,un1
n1
X
r=0
n1
X
r=0
Using the properties of the single entry matrix found in Sec. 9.7.4, the result
follows easily.
B.1.2
det(XH AX)
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 67
B.1
Misc Proofs
Through the calculations, (92) and (219) were used. In addition, by use of (220),
the derivative is found with respect to the imaginary part of X
i
det(XH AX)
=X
=
=
1 det(XH AX)
det(XH AX)
i
2
<X
=X
T
H
H
1 H
det(X AX) (X AX) X A
=
=
1 det(XH AX)
det(XH AX)
+i
2
<X
=X
H
H
1
det(X AX)AX(X AX)
Notice, for real X, A, the sum of (228) and (229) is reduced to (48).
Similar calculations yield
det(XAXH )
X
=
=
1 det(XAXH )
det(XAXH )
i
2
<X
=X
H
H
H 1 T
det(XAX ) AX (XAX )
(545)
1 det(XAXH )
det(XAXH )
+i
2
<X
=X
det(XAXH )(XAXH )1 XA
(546)
and
det(XAXH )
X
=
=
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 68
REFERENCES
REFERENCES
References
[1] Karl Gustav Andersson and Lars-Christer Boiers. Ordinaera differentialekvationer. Studenterlitteratur, 1992.
[2] J
orn Anem
uller, Terrence J. Sejnowski, and Scott Makeig. Complex independent component analysis of frequency-domain electroencephalographic
data. Neural Networks, 16(9):13111323, November 2003.
[3] S. Barnet. Matrices. Methods and Applications. Oxford Applied Mathematics and Computin Science Series. Clarendon Press, 1990.
[4] Christopher Bishop. Neural Networks for Pattern Recognition. Oxford
University Press, 1995.
[5] Robert J. Boik. Lecture notes: Statistics 550. Online, April 22 2002. Notes.
[6] D. H. Brandwood. A complex gradient operator and its application in
adaptive array theory. IEE Proceedings, 130(1):1116, February 1983. PTS.
F and H.
[7] M. Brookes. Matrix Reference Manual, 2004. Website May 20, 2004.
[8] Contradsen K., En introduktion til statistik, IMM lecture notes, 1984.
[9] Mads Dyrholm. Some matrix results, 2004. Website August 23, 2004.
[10] Nielsen F. A., Formula, Neuro Research Unit and Technical university of
Denmark, 2002.
[11] Gelman A. B., J. S. Carlin, H. S. Stern, D. B. Rubin, Bayesian Data
Analysis, Chapman and Hall / CRC, 1995.
[12] Gene H. Golub and Charles F. van Loan. Matrix Computations. The Johns
Hopkins University Press, Baltimore, 3rd edition, 1996.
[13] Robert M. Gray. Toeplitz and circulant matrices: A review. Technical
report, Information Systems Laboratory, Department of Electrical Engineering,Stanford University, Stanford, California 94305, August 2002.
[14] Simon Haykin. Adaptive Filter Theory. Prentice Hall, Upper Saddle River,
NJ, 4th edition, 2002.
[15] Roger A. Horn and Charles R. Johnson. Matrix Analysis. Cambridge
University Press, 1985.
[16] Mardia K. V., J.T. Kent and J.M. Bibby, Multivariate Analysis, Academic
Press Ltd., 1979.
[17] Mathpages on Eigenvalue Problems and Matrix Invariants,
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 69
REFERENCES
REFERENCES
http://www.mathpages.com/home/kmath128.htm
[18] Carl D. Meyer. Generalized inversion of modified matrices. SIAM Journal
of Applied Mathematics, 24(3):315323, May 1973.
[19] Thomas P. Minka. Old and new matrix algebra useful for statistics, December 2000. Notes.
[20] Daniele Mortari OrthoSkew and OrthoSym Matrix Trigonometry John
Lee Junkins Astrodynamics Symposium, AAS 03265, May 2003. Texas
A&M University, College Station, TX
[21] L. Parra and C. Spence. Convolutive blind separation of non-stationary
sources. In IEEE Transactions Speech and Audio Processing, pages 320
327, May 2000.
[22] Kaare Brandt Petersen, Jiucang Hao, and Te-Won Lee. Generative and
filtering approaches for overcomplete representations. Neural Information
Processing - Letters and Reviews, vol. 8(1), 2005.
[23] John G. Proakis and Dimitris G. Manolakis. Digital Signal Processing.
Prentice-Hall, 1996.
[24] Laurent Schwartz. Cours dAnalyse, volume II. Hermann, Paris, 1967. As
referenced in [14].
[25] Shayle R. Searle. Matrix Algebra Useful for Statistics. John Wiley and
Sons, 1982.
[26] G. Seber and A. Lee. Linear Regression Analysis. John Wiley and Sons,
2002.
[27] S. M. Selby. Standard Mathematical Tables. CRC Press, 1974.
[28] Inna Stainvas. Matrix algebra in differential calculus. Neural Computing
Research Group, Information Engeneering, Aston University, UK, August
2002. Notes.
[29] P. P. Vaidyanathan. Multirate Systems and Filter Banks. Prentice Hall,
1993.
[30] Max Welling. The Kalman Filter. Lecture Note.
[31] Wikipedia on minors: Minor (linear algebra),
http://en.wikipedia.org/wiki/Minor_(linear_algebra)
[32] Zhaoshui He, Shengli Xie, et al, Convolutive blind source separation in
frequency domain based on sparse representation, IEEE Transactions on
Audio, Speech and Language Processing, vol.15(5):1551-1563, July 2007.
Petersen & Pedersen, The Matrix Cookbook, Version: November 14, 2008, Page 70
Index
Anti-symmetric, 49
Block matrix, 41
Chain rule, 14
Cholesky-decomposition, 28
Co-kurtosis, 29
Co-skewness, 29
Condition number, 61
Cramers Rule, 57
Derivative of a complex matrix, 22
Derivative of a determinant, 7
Derivative of a trace, 11
Derivative of an inverse, 8
Derivative of symmetric matrix, 15
Derivatives of Toeplitz matrix, 15
Dirichlet distribution, 33
Eigenvalues, 26
Eigenvectors, 26
Exponential Matrix Function, 55
Nilpotent, 44
Norm of a matrix, 59
Norm of a vector, 59
Normal-Inverse Gamma distribution, 33
Normal-Inverse Wishart distribution, 34
Orthogonal, 44
Power series of matrices, 54
Probability matrix, 51
Pseudo-inverse, 20
Schur complement, 36, 42
Single entry matrix, 47
Singular Valued Decomposition (SVD),
26
Skew-Hermitian, 43
Skew-symmetric, 49
Stochastic matrix, 51
Student-t, 32
Sylvesters Inequality, 61
Symmetric, 49
Gaussian, conditional, 36
Gaussian, entropy, 40
Gaussian, linear combination, 36
Gaussian, marginal, 35
Gaussian, product of densities, 37
Generalized inverse, 20
Taylor expansion, 54
Toeplitz matrix, 50
Transition matrix, 51
Trigonometric functions, 55
Hermitian, 43
Vandermonde matrix, 53
Vec operator, 55, 56
Unipotent, 44
Idempotent, 44
Kronecker product, 55
Wishart distribution, 33
Woodbury identity, 17
LDL decomposition, 28
LDM-decomposition, 28
Linear regression, 56
LU decomposition, 28
Lyapunov Equation, 59
Moore-Penrose inverse, 20
Multinomial distribution, 33
71