Homework #3 - Problem 2

[Rajpreet Thethy]

Hi All,

I have a question about problem #2 for Homework #3. I'm having a hard time setting up with model. It seems my A / (Yt-1) function is not coming out correctly. Can someone please review my attached Excel file and see if they can figure out what I am doing wrong.

Thanks!
Rajpreet

[Nicholas Woo]

You don't need to calculate cumulative sales. Set Yt equal to the given sales data and you'll be fine.

[Rajpreet Thethy]

Hi Nicholas - Even when I do that the A / (Yt-1) function does not seem correct. Please see attached.

Thanks for your help.
Rajpreet

[Nicholas Woo]

Also, you may want to change your value for A. You currently have it set to 90, but that is suggesting you expect sales to level out at 90 customers out of the total 150,000 possible.

On Sat, Apr 9, 2011 at 6:40 PM, Rajpreet Thethy <rajp...@gmail.com> wrote:

[Nicholas Woo]

1.) I would set A to something larger than 90. I actually used 75,000, but this number should represent the number of customers you expect sales to level off at within the next 10 years.

2.) You should be graphing the LN function of A/Yt-1, which will then be your linearized actual sales.

3.) Run a regression on your results, and use them to create your forecast column.

[Rajpreet Thethy]

Thanks Nicholas. I see what I was doing wrong. I thought A was a percentage. That's why I had set it to 90%. Thanks for the help.

Rajpreet

[Samuel Harrison]

Also - watch how you set up your equation.
It is supposed to be: (A/Yt) - 1
Which is entirely different than A/(Yt-1).
And then: ln((A/Yt)-1)

Sam Harrison

[Ajay Goyal]

attached is my solution.

Does it match with any of you guys?

[Ashish Swaroop]

Ajay,

Shouldn't you use C and B values from regression output?

rest looks good to me.

regds

[Ajay Goyal]

Hi Ashish,

I used the following regression output

C=exp(2.567036) 

B=-0.13897

 

is this what you used? or something else?

	Coefficients	Standard Error	t Stat	P-value	Lower 95%	Upper 95%	Lower 95.0%	Upper 95.0%
Intercept	2.567306	0.071684	35.81407	2.23E-14	2.412442	2.722171	2.412442	2.722171
year	-0.13897	0.007884	-17.6265	1.86E-10	-0.156	-0.12194	-0.156	-0.12194

 

 

thanks

ajay

	Coefficients
Intercept	2.567306
year	-0.13897

[Crystal]

Ajay,

Why did you choose to use 135,000 as A? Is it because 135,000 is 90%
of 150,000?

Crystal

[Ajay Goyal]

Crystal,

yes. We had a group session on Saturday and we decided to use 90%.

What are you  using?

 

thanks

ajay

[Crystal]

Ajay,

I missed that session. I used 150,000 since these are their existing
customer and I figured they would also gain new customers over the
next ten years. I think I might redo my work to reflect the 90%
though.

-Crystal

[Xiuya Li]

Any one has result for HW3 problem 1 yet? log method works better than linear, but none of the R^2 is more than 0.2, which is useless in statistics.

Ted

[Samuel Harrison]

Which is why the answer - for me - on the question regarding the potential validity of the forecast is that it is, more likely than not, wrong.

[Kartik Ramachandran]

My answers for Hw 4..let me know if you get anything different.

MA3    

MAD 55.44

MSE 5,522.66426180558 

MAPE 0.490406984 

MA7           

MAD 81.47285714 

MSE 9940.603038  

MAPE 0.721155834 

ES .5   

MAD 49.47165226 

MSE 4600.868401 

MAPE 0.437643252  

Thanks,

Kartik