Professional Documents
Culture Documents
• Linear regression is a model that uses least squares method to identify the relationship
between an independent variable and one or more dependent variables.
• In simple regression there is one independent variable and in multiple regression there
are more than one independent variables.
• In simple regression the dependent variable is a function of one independent variable
and therefore the theoretical relationship is the straight line.
Where
Also
Co –efficient of correlation
1) Café coffee produces commercial coffee cups that are sold all over the world. The company is
operating at near full capacity for over a year. Plant manager thinks the sales growth will
continue and he wants to develop long range forecast to help the plant facility requirement for
the next 3 years. Sales records for the past 10 years have been compiled.
Year 1 2 3 4 5 6 7 8 9 10
Annual sales 1000 1300 1800 2000 2000 2000 2200 2600 2900 3200
(000’s of units)
Solution:
X -> Year ; Y-> Annual Sales (000’s of units)
X Y XY X2
1 1,000 1,000 1
2 1,300 2,600 4
3 1,800 5,400 9
4 2,000 8,000 16
5 2,000 10,000 25
6 2,000 12,000 36
7 2,200 15,400 49
8 2,600 20,800 64
9 2,900 26,100 81
10 3,200 32,000 100
55 21,000 1,33,300 385
Y = a+bx
For 11, 12, 13 years the forecasted values are 3287, 3502 & 3718 respectively.
2) Jack the general manager of precession engineering corporation thinks that his firms
engineering services supplied to high-way construction firm are directly related to the amount
of highway construction contracts released in his geographic area. Jack asked Mr.Malya one of
his engineers to perform a simple linear regression analysis on historical data.
a) Develop a regression equation for predicting the level of demand.
b) Use the regression equation to predict the level of demand for the next four quarters for
which the amount of contracts released are 260, 290, 300 & 270 (thousands)
c) Determine how closely demand is related to the amount of construction released.
Solution:
Linear regression equation; Y= a+bX
Y- Sales; X- Total amt of contract released.
Y X XY Y2 X2
8 150 1200 64 22500
10 170 1700 100 28900
15 190 2850 225 36100
9 170 1530 81 28900
12 180 2160 144 32400
13 190 2470 169 36100
12 200 2400 144 40000
16 220 3520 256 48400
95 1470 17830 1183 273300
In problems; r2=0.79 – 80%. This means that amount of contracts released (x) explain 80% of
variation in sales of precision instruments Remaining 20% of the variation can be attributed to
other causal factors.
Solution:
X-Weeks; Y-Sales.
X Y XY X2
1 700 700 1
2 724 1448 4
3 720 2160 9
4 728 2912 16
5 740 3700 25
6 742 4452 36
7 758 5306 49
8 750 6000 64
9 770 6930 81
10 775 7750 100
55 7407 41358 385
Y = a+bx
Y = a+bx
r= 0.901; this means that there is strong & positive relationship between sales & advertising
expenditure.
r2 = 0.8125; this means that the variation in the sales can be attributed to an extent of 81.25% to
the variation in advertising expenditure.
X Y XY X2
200 165 33,000 40,000
235 184 43,240 55,225
210 180 37,800 44,100
197 145 28,565 38,809
225 190 42,750 50,625
240 169 40,560 57,600
217 180 39,060 47,089
225 170 38,250 50,625
1749 1383 3,03,225 3,84,073
Y = a+bx