Korrelyatsion-regression tahlilga kiritilgan omillar qiymatlari
Yillar
x
1
x
2
x
3
x
4
x
5
x
6
y
2004
295,3
16
30 025 660,0 26 600 452,0
2 560,0 12 505,5
8 456,2
2005
367,9
16
32 880 200,0 28 215 400,0
2 890,0 13 598,7
9 006,3
2006
489,1
16
36 254 820,0 31 156 880,0
3 090,0 14 980,9
9 450,5
2007
621,2
14
38 255 015,0 33 550 400,0
3 569,0 15 875,2
9 880,2
2008
838,5
14
40 258 700,0 35 800 026,0
4 215,0 16 020,6 10 564,2
2009 1060,7 14
49 256 860,0 38 256 900,0
4 890,0 16 900,7 10 985,3
2010 1405,5 14
53 596 700,0 42 501 000,0
5 450,0 17 899,7 11 348,7
2011 1730,2 12
24 479 000,0
792 100,0
6 730,0 25 650,0 14 911,7
2012 2601,9 12
40 671 500,0
1 450 600,0 15 886,2 28 277,6 14 710,5
2013 3166,5 12
45 860,8
31 217,7 41 963,5 18 496,0
74,5
2014 3489,5 10
36 818,0
37 323,5 52 935,0 19 701,5
149,1
Omillar bog’liqligining ekonometrik modelini aniqlash uchun ko’p omilli
korelyatsion-regression tahlil usulidan foydalaniladi. Ishda samaradorlik
ko’rsatkichlarini tahlil qilish uchun quyidagi ekonometrik modellardan (ko’p
omilli regressiya tenglamalardan) foydalanildi:
1)
i
m
i
i
х
y
∑
=
+
=
1
0
β
β
(1) Chiziqli model ;
2)
∑
=
+
=
m
i
i
i
x
y
1
0
ln
β
β
(2) Logorifmik model;
3)
∑
=
+
=
m
i
i
i
x
y
1
0
β
β
(3) Giperbolik model;
103
4)
∏
=
=
m
i
i
i
x
y
1
0
β
β
(4) Ko’rsatkichli model.
Bunda,
β
0
– ozod had;
y – shirkat sof foydasi ko’rsatkichisi;
x
i
– shirkat sof foydasiga ta’sir etuvchi omillar;
β
i
– ko’p omilli model parametrlari; (i= 1,2,3....m);
m – tanlangan omillar soni.
Berilgan y = f (x
1
, x
2
, .... , x
n
) bog’liqlikni topish zarur. Ushbu bog’liqlikni
aniqlashda eng kichik kvadratlar usulidan foydalanamiz.
26-jadval
Ko’p omilli regression tahlilda foydalanilgan belgilar
№
Nomi
Ifodalanishi
1
Konstanta, ozod had
β
0
2
Ko’p omilli korrelyatsiya
koeffitsienti
R
2
3
Ko’p omilli regressiya
tenglamasining noma’lum
parametrlari
Β
1
, β
2
,... β
k
4
Baholashning standart xatosi
Qiymati qancha kichik bo’lsa model
shuncha ahamiyatli hisoblanadi
5
Determinatsiya koeffitsienti
Qiymati qanchalik 1 sonida yaqin
model adekvat hisoblanadi
6
F – Fisher koeffitsienti
F
haq
> F
tabl
bo’lsa determinatsiya
koeffitsienti ahamiyatli hisoblanadi
7
P-qiymat
(“nolli” gipotezaning
ahamiyatliligini ko’rsatuvchi mezon)
0,05 dan kichik ko’rsatkich tegishli
model ahamiyatli hisoblanadi
Ko’p omilli korrelyatsion bog’lanishning xususiyati shundaki, uning
regressiya tenglamasida bir necha muhim va mohiyatli omillar ishtirok etadi. Bu
omillardan eng mohiyatlisini to’g’ri tanlash va ularni regressiya tenglamasiga
kiritish katta ahamiyatga egadir. Omillarni tanlash va sifat jihatdan nazariy tahlil
qilishga asoslanadi va uch bosqichda o’tkaziladi. Birinchi bosqichda (dastlabki
tahlilda) omillar hech qanday shart qo’yilmasdan tanlanadi. Ikkinchi bosqichda
104
ular juft korrelyatsiya koeffitsentlaridan foydalangan holda tahlil qilinadi.
Buning uchun belgilar y
Do'stlaringiz bilan baham: |