Embed
Email

Matrix Algebra

Document Sample
Matrix Algebra
Shared by: HC111126103739
Categories
Tags
Stats
views:
1
posted:
11/26/2011
language:
English
pages:
45
Matrix Algebra

 Matrix algebra is a means of expressing

large numbers of calculations made

upon ordered sets of numbers.

 Often referred to as Linear Algebra

Why use it?

 Matrix algebra is used primarily to

facilitate mathematical expression.

 Many equations would be completely

intractable if scalar mathematics had to

be used. It is also important to note

that the scalar algebra is under there

somewhere.

Definitions - scalar

 scalar - a number

 denoted with regular type as is scalar

algebra

 [1] or [a]

Definitions - vector

 vector - a single row or column of

numbers

 denoted with bold small letters

 row vector 1 2 3 4 5

1 

 column vector 2

 

3

 

4

5 

 

Definitions - Matrix

 A matrix is a set of rows and columns of

numbers

1 2 3

 4 5 6

 





 Denoted with a bold Capital letter

 All matrices (and vectors) have an order

- that is the number of rows x the

number of columns. Thus A is

1 2 3

 4 5 6

  2 x 3 

Matrix Equality

 Thus two matrices are equal iff (if and

only if) all of their elements are

identical

 Note: your data set is a matrix.

Matrix Operations

 Addition and Subtraction

 Multiplication

 Transposition

 Inversion

Addition and Subtraction

 Two matrices may be added iff they are

the same order.

 Simply add the corresponding elements



a11 a12   b11 b12  c11 c12 

a a22   b21 b22   c 21 c 22 

 21     

a31 a32  b31 b32  c31 c32 

     

Addition and Subtraction

(cont.)

 Where

a11  b11  c11

a12  b12  c12

a21  b21  c 21

a22  b22  c 22

a31  b31  c31

a32  b32  c32



 Hence

1 2 4 6 5 8 

3 4  4 6  7 10

     

5 6 4 6 9 12

     

Matrix Multiplication

 To multiply a scalar times a matrix,

simply multiply each element of the

matrix by the scalar quantity



a11 a12   2a11 2a12 

2   2a

a21 a22   21 2a22 

 

Matrix Multiplication (cont.)

 To multiply a matrix times a matrix, we

write

 A times B as AB

 This is pre-multiplying B by A, or post-

multiplying A by B.

Matrix Multiplication (cont.)

 In order to multiply matrices, they must

be conformable (the number of columns

in A must equal the number of rows in

B.)

 an (mxn) x (nxp) = (mxp)

 an (mxn) x (pxn) = cannot be done

 a (1xn) x (nx1) = a scalar (1x1)

Matrix Multiplication (cont.)

 Thus a11 a12 a13   b11 b12  c11 c12 

a a22 a23  x b21 b22   c 21 c 22 

 21     

a31 a32 a33  b31 b32  c31 c32 

     



 where

c11  a11b11  a12 b21  a13b31

c12  a11b12  a12 b22  a13b32

c 21  a21b11  a22 b21  a23b31

c 22  a21b12  a22 b22  a23b32

c31  a31b11  a32 b21  a33b31

c32  a31b12  a32 b22  a33b32

Matrix Multiplication- an

example

 Thus 1 4 7 1 4 c11 c12  30 66

2 5 8 x 2 5  c c 22   36 81

     21   

3 6 9 3 6 c31 c32  42 96

       



 where

c11  1 * 1  4 * 2  7 * 3  30

c12  1 * 4  4 * 5  7 * 6  66

c 21  2 * 1  5 * 2  8 * 3  36

c 22  2 * 4  5 * 5  8 * 6  81

c31  3 * 1  6 * 2  9 * 3  42

c32  3 * 4  6 * 5  9 * 6  96

Matrix multiplication is not

Commutative

 AB does not necessarily equal BA

 (BA may even be an impossible

operation)

Yet matrix multiplication is

Associative

 A(BC) = (AB)C

Special matrices

 There are a number of special matrices

 Diagonal

 Null

 Identity

Diagonal Matrices

 A diagonal matrix is a square matrix that

has values on the diagonal with all off-

diagonal entities being zero.



a11 0 0 0

0 a 0 0

 22 

0 0 a33 0

 

0 0 0 a44 

Identity Matrix

 An identity matrix is a diagonal matrix

where the diagonal elements all equal

one. It is used in a fashion analogous

to multiplying through by "1" in scalar

math.

1 0 0 0

0 1 0 0

 

0 0 1 0

 

0 0 0 1

Null Matrix

 A square matrix where all elements equal

zero.

0 0 0 0

0 0 0 0

 

0 0 0 0

 

0 0 0 0

 Not usually „used‟ so much as sometimes the

result of a calculation.

 Analogous to “a+b=0”

The Transpose of a Matrix A'

 Taking the transpose is an operation

that creates a new matrix based on an

existing one.

 The rows of A = the columns of A'

 Hold upper left and lower right corners

and rotate 180 degrees.

Example of a transpose





1 4 

2 5, A'  1 2 3

A   4 5 6

3 6  

 

The Transpose of a Matrix A'

 If A = A', then A is symmetric (i.e. correlation

matrix)

 If AA‟ = A then A' is idempotent (and A' = A)

 The transpose of a sum = sum of transposes

( A  B  C )'  A'B'C'

 The transpose of a product = the product of

the transposes in reverse order

( ABC )'  C' B' A'

An example:

 Suppose that you wish to obtain the

sum of squared errors from the vector

e. Simply pre-multiply e by its

transpose e'.

 which, in matrices looks like





e' e  e1  e2  ..en

2 2 2



An example - cont

 Since the matrix product is a scalar

found by summing the elements of the

vector squared.

The Determinant of a Matrix

 The determinant of a matrix A is

denoted by |A|.

 Determinants exist only for square

matrices.

 They are a matrix characteristic, and

they are also difficult to compute

The Determinant for a 2x2 matrix

 If A = a11 a12 

a 

 21 a22 









 Then

A  a11a22  a12a21





 That one is easy

The Determinant for a 3x3 matrix

 If A =

a11 a12 a13 

a 

 21 a22 a23 

a31 a32

 a33 





 Then



A  a11a22a33  a11a23a32  a12a23a31  a12a21a33  a13a21a32  a13a22a31

Determinants

 For 4 x 4 and up don't try. For those

interested, expansion by minors and

cofactors is the preferred method.

 (However the spaghetti method works

well! Simply duplicate all but the last

column of the matrix next to the

original and sum the products of the

diagonals along the following pattern.)

Spaghetti Method of |A|





a11 a12 a13  a11 a12 

a a22  a 

a23   21 a22 

 21



a31 a32

 a33  a31 a32 

 

Properties of Determinates

 Determinants have several mathematical

properties which are useful in matrix

manipulations.

 1 |A|=|A'|.

 2. If a row of A = 0, then |A|= 0.

 3. If every value in a row is multiplied by k,

then |A| = k|A|.

 4. If two rows (or columns) are interchanged

the sign, but not value, of |A| changes.

 5. If two rows are identical, |A| = 0.

Properties of Determinates

 6. |A| remains unchanged if each

element of a row or each element

multiplied by a constant, is added to any

other row.

 7. Det of product = product of Det's

|A| = |A| |B|

 8. Det of a diagonal matrix = product

of the diagonal elements

The Inverse of a Matrix (A-1)

 For an nxn matrix A, there may be a B such

that AB = I = BA.

 The inverse is analogous to a reciprocal)

 A matrix which has an inverse is nonsingular.

 A matrix which does not have an inverse is

singular.

 An inverse exists only if A  0

Inverse by Row or column

operations

 Set up a tableau matrix

 A tableau for inversions consists of the

matrix to be inverted post multiplied by

a conformable identity matrix.

Matrix Inversion by Tableau

Method

 Rules:

 You may interchange rows.

 You may multiply a row by a scalar.

 You may replace a row with the sum of that row

and another row multiplied by a scalar.

 Every operation performed on A must be

performed on I

 When you are done; A = I & I = A-1

The Tableau Method of Matrix

Inversion: An Example

 Step 1: Set up Tableau



1 4 3  1 0 0

2 5 4  0 1 0 

  

1  3  2  0 0 1 

  

Matrix Inversion – cont.

 Step 2: Add –2(Row 1) to Row 2

1 4 3   1 0 0

0  3  2    2 1 0 

  

1  3  2  0 0 1

  



 Step 3: Add –1(Row 1) to Row 3

1 4 3   1 0 0

0  3  2    2 1 0 

  

0  7  5    1 0 1 

  

Matrix Inversion – cont.

 Step 4: Multiply Row 2 by –1/3

1 4 3  1 0 0

0 1 2 / 3 2 / 3  1 / 3 0

  

0  7  5    1

  0 1





 Step 5: Add –4 (Row 2) to Row 1

1 0 1 / 3   5 / 3 4 / 3 0

0 1 2 / 3  2 / 3  1 / 3 0

  

0  7  5    1

  0 1



Matrix Inversion – cont.

 Step 6: Add 7(Row 2) to Row 3

1 0 1 / 3    5 / 3 4 / 3 0

0 1 2 / 3   2 / 3  1 / 3 0

  

0 0  1 / 3  11 / 3  7 / 3 1

  



 Step 7: Add Row 3 to Row 1

1 0 0  2  1 1

0 1 2 / 3   2 / 3  1 / 3 0 

  

0 0  1 / 3 11 / 3  7 / 3 1

  

Matrix Inversion – cont.

 Step 9: Add 2(Row 3) to Row 2

1 0 0  2  1 1

0 1 0  8  5 2

  

0 0  1 / 3 11 / 3  7 / 3 1

  



 Step 9: Multiply Row 3 by -3

1 0 0   2 1 1 

0 1 0   8  5 2 

  

0 0 1  11 7  3

  

Checking the calculation

 Remember AA-1=I

1 4 3  2  1 1  1 0 0

2 5 4   8  5 2   0 1 0 

    

1  3  2  11 7  3 0 0 1

    







 Thus

1 * 2  4 * 8  3 * 11  1

1 * 1  4 * 5  3 * 7  0

etc

The Matrix Model

 The multiple regression model may be

easily represented in matrix terms.

Y  XB  e



 Where the Y, X, B and e are all matrices

of data, coefficients, or residuals

The Matrix Model (cont.)

 The matrices in Y  XB  e are

represented by



 Y1   X 11 X 12 ... X ik   B1   e1 

Y  X X ... X 2 k  B  e 

Y  

2

X  21 22  B  

2

e  

2



  ... ... ... ...     

       

 Yn   X n1 X n 2 ... X nk   Bk   en 



 Note that we postmultiply X by B since this

order makes them conformable.

The Assumptions of the Model

Scalar Version

 1. The ei's are normally distributed.

 2. E(ei) = 0

 3. E(ei2) = 2

 4. E(eiej) = 0 (ij)

 5. X's are nonstochastic with values fixed in repeated

samples and (Xik-Xbark)2/n is a finite nonzero

number.

 6. The number of observations is greater than the

number of coefficients estimated.

 7. No exact linear relationship exists between any of

the explanatory variables.

The Assumptions of the

Model: The Matrix Version

 These same assumptions expressed in

matrix format are:



 1. e  N(0,)

 2.  = 2I

 3. The elements of X are fixed in repeated

samples and (1/ n)X'X is nonsingular and

its elements are finite


Related docs
Other docs by HC111126103739
2nd -1er
Views: 2  |  Downloads: 0
Decreto 006 de 1998
Views: 1  |  Downloads: 0
Ma�tre d�ouvrage :
Views: 3  |  Downloads: 0
tributario
Views: 11  |  Downloads: 0
PUISSANCE DE DIX
Views: 5  |  Downloads: 0
Proposition de th�se
Views: 0  |  Downloads: 0
seance_1_etat_art_commerce_electronique
Views: 1  |  Downloads: 0
By registering with docstoc.com you agree to our
privacy policy

You are almost ready to download!

You are almost ready to download!