,WEEKl
190 B1) E (ly-bo b)
angmin
=
,
+
B
,
CoVly x)
=
,
van (l)
Bo E(x)
Ely)
=
-
B,
ASSUME E(u) =
0
E(n(x) =
Elu) =
o
TOESTIMATE Bo , B1:
E(U) =
0
Cov(x, n) =
0
0
Elsu) -
EGDE(n)
=
E(U) =
LAW OF ITERATIVE ECE(xuIX)) ECEluDH)
EXPECTATION :
ECU) =
= =
0
METHODS OF MOMENTS
E(U) =
Ely-Bo-Bix) = 0
Elu) =
Elly-Bo-Bix]] = 0
,① Elyi -
50 -
Bx) =
0 -
y
-
3 -
Bx 0E) y - B x
= =
② [silyi-5 Bx) -
=
0
#)
[xi(yi -
15 B)
-
-
Spi) = 0
27
Exillyi y) -
+ &(c xi)) -
=
0
()( =
[xi(yi j) -
, zxki -) + 0
=(i(x -
x)
& xi
-i)(yc y) -
=
[xilyiy)
Epi -zi)(yi-y)
& =
= covk , yi)
= (2 -e) 2 Cor( <(i)
ORDINARY LEAST SQUARES
(50 5)
any min Ilyi-Bo-Bis-
=
,
sameResults -
-
p=
↑ =
Exilyi) = Ex-)(yi y) -
"
Exibli-ci) Ecosi)
, [m = 0
Zim =0
5 =
3 + (, j)
always on OLs
SST = lyi-y)2
SSE-Zy-y) SSR =
Zi
SST =
[ (lyi yf) (y y))
-
+ -
=
z(m +
( -
j)) zym
=0
SST =
SSR + SSE
RE
190 B1) E (ly-bo b)
angmin
=
,
+
B
,
CoVly x)
=
,
van (l)
Bo E(x)
Ely)
=
-
B,
ASSUME E(u) =
0
E(n(x) =
Elu) =
o
TOESTIMATE Bo , B1:
E(U) =
0
Cov(x, n) =
0
0
Elsu) -
EGDE(n)
=
E(U) =
LAW OF ITERATIVE ECE(xuIX)) ECEluDH)
EXPECTATION :
ECU) =
= =
0
METHODS OF MOMENTS
E(U) =
Ely-Bo-Bix) = 0
Elu) =
Elly-Bo-Bix]] = 0
,① Elyi -
50 -
Bx) =
0 -
y
-
3 -
Bx 0E) y - B x
= =
② [silyi-5 Bx) -
=
0
#)
[xi(yi -
15 B)
-
-
Spi) = 0
27
Exillyi y) -
+ &(c xi)) -
=
0
()( =
[xi(yi j) -
, zxki -) + 0
=(i(x -
x)
& xi
-i)(yc y) -
=
[xilyiy)
Epi -zi)(yi-y)
& =
= covk , yi)
= (2 -e) 2 Cor( <(i)
ORDINARY LEAST SQUARES
(50 5)
any min Ilyi-Bo-Bis-
=
,
sameResults -
-
p=
↑ =
Exilyi) = Ex-)(yi y) -
"
Exibli-ci) Ecosi)
, [m = 0
Zim =0
5 =
3 + (, j)
always on OLs
SST = lyi-y)2
SSE-Zy-y) SSR =
Zi
SST =
[ (lyi yf) (y y))
-
+ -
=
z(m +
( -
j)) zym
=0
SST =
SSR + SSE
RE
