In: Statistics and Probability
The following table of values gives a company's annual profits
in millions of dollars. Rescale the data so that the year 2005
corresponds to x=1 .
Year | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 |
Profits (in millions of dollars) | 51.8 | 63.4 | 66.3 | 65.8 | 62.1 | 63.8 |
Use the power regression model to predict the company's profits in
2012.
a) $70.2 million
b) $68.6 million
c) $71.5 million
d) $67.7 million
e) $66.6 million
f) None of the above
we have
power regression ----------
Y =a(x)b
ln(y) = ln(a) + b*ln(x)
now, this is equivalent to linear equation Y= a + bx
x | y | ln(x) | ln(y) |
1 | 51.8 | 0 | 3.94739 |
2 | 63.4 | 0.693147 | 4.149464 |
3 | 66.3 | 1.098612 | 4.19419 |
4 | 65.8 | 1.386294 | 4.18662 |
5 | 62.1 | 1.609438 | 4.128746 |
6 | 63.8 | 1.791759 | 4.155753 |
(x-x̅)² | (y-ȳ)² | (x-x̅)(y-ȳ) |
1.20 | 0.03 | 0.20 |
0.16 | 0.00 | -0.01 |
0.00 | 0.00 | 0.00 |
0.08 | 0.00 | 0.02 |
0.26 | 0.00 | 0.00 |
0.48 | 0.00 | 0.02 |
ΣX | ΣY | Σ(x-x̅)² | Σ(y-ȳ)² | Σ(x-x̅)(y-ȳ) | |
total sum | 6.579251212 | 24.76216293 | 2.195482001 | 0.0 | 0.23 |
mean | 1.10 | 4.13 | SSxx | SSyy | SSxy |
sample size , n = 6
here, x̅ = Σx / n= 1.10 ,
ȳ = Σy/n = 4.13
SSxx = Σ(x-x̅)² = 2.1955
SSxy= Σ(x-x̅)(y-ȳ) = 0.2
estimated slope , ß1 = SSxy/SSxx = 0.2
/ 2.195 = 0.1030
intercept, ß0 = y̅-ß1* x̄ =
4.0141
so, regression line is Ŷ =
4.0141 + 0.1030 *x
...............
for year 2012 x= 8
ln(8) = 2.07944
Predicted Y at X= 2.079441542
is
Ŷ = 4.01406 +
0.103024 * 2.079441542
= 4.228
now,
company's profit = antilog(4.228) = 68.6 ( excel formula =EXP(4.228) )
so answer is option (b) $68.6
million