Topic 2 The General Linear Model - Powerpoint PDF
Topic 2 The General Linear Model - Powerpoint PDF
Econometrics:
Topic 2 The General Linear Model
Kevin Albertson
[email protected]
MMUBS 4·12
Office Hours: Tuesday and Thursday at 12:00
Summary so far
3
Manchester Metropolitan University
4
Matrix revision
Matrix revision
5
Matrix revision
Matrix multiplication
b1 j
b2 j
ai1 ai 2 ai 3 ain b3 j cij
bnj
6
Matrix revision
The transpose of a matrix
2 3
- 1 1
2 - 1 7 5
A A
3 1 0 7 0
5
7
Matrix revision
The inverse of a matrix
(AB)’ = B’A’
(AB)⁻¹ = B⁻¹A⁻¹
9
Matrix revision
Matrix rank
12
Matrix revision
Matrix rank
1 1 1
_2 1 3_has a rank of 3
5 1 5
13
Matrix revision
Properties of matrix rank
14
Random vectors
x E x x
y E y
EL E y
z E z
z
16
Random vectors
2x x, y x,z
x,y y y ,z
2
L
2
x , z y , z z
17
Random vectors
21
The general linear model
Matrix notation
y₁ = [ 1 x₁ w₁ … z₁] + ε₁
0
y₂ = [ 1 x₂ w₂ … z₂] + ε₂
y₃ = [ 1 x₃ w₃ … z₃] 1 + ε₃
:
yn = [ 1 xn wn … zn]
K 1 + εn
22
The general linear model
Matrix notation
y₁ = [ 1 x₁ w₁ … z₁] + ε₁
0
y₂ = [ 1 x₂ w₂ … z₂] + ε₂
y₃ = [ 1 x₃ w₃ … z₃] 1 + ε₃
:
yn = [ 1 xn wn … zn]
K 1 + εn
23
The general linear model
Matrix notation
y1 1 x1 w1 z1 1
y 1 0
x2 w2 z2
2 2 2
y3 1 x3 w3 z3 3
K 1
yn 1 xn wn z n n
y = Xβ + ε
24
The general linear model
Matrix notation
33
The general linear model
Properties of estimates
34
Manchester Metropolitan University
The variance and standard error of the OLS estimator
35
The general linear model
Properties of estimates
• As b = (X’X)⁻¹X’y
b ~ N(β, σ²(X’X)⁻¹)
SSR
• An estimator of var(εᵢ) = σ² is s
2
nK
36
The general linear model
Properties of estimates
• Measures of Variation
n n n
SST yi y 2
SSE yˆi y 2 SSR yi yˆi 2
i 1 i 1 i 1
40
The general linear model
R-Squared – R²
41
Manchester Metropolitan University
42
Summary
Summary