| |
 |
|
|
Science Forum Index » Statistics - Math Forum » What statistical method to use to predict sales per...
Page 1 of 1
|
| Author |
Message |
| sailor... |
 Posted: Wed Jul 02, 2008 10:32 am |
|
|
|
Guest
|
All,
I would like to predict the sales performance (# of articles sold, #
of articles required) based on the sales data of each store (# of
articles distributed, # of articles sold, regular product, and so on),
but it seems like I don't find my own way in the jungle of statistics.
What would be the best model to use? Should this just be multivariate
regression? I have seen so many terms like 'time series analysis', '
regression analysis' 'neural network', and am not sure what model to
use.
In case you cannot suggest any model, could you tell me how I should
decide what model to use?
thanks in advance,
Cornelis |
|
|
| Back to top |
|
| Stephen Wolstenholme... |
| Posted: Wed Jul 02, 2008 5:06 pm |
|
|
|
Guest
|
On Wed, 2 Jul 2008 13:32:20 -0700 (PDT), sailor <BJCornelis at (no spam) gmail.com>
wrote:
Quote: What would be the best model to use? Should this just be multivariate
regression? I have seen so many terms like 'time series analysis', '
regression analysis' 'neural network', and am not sure what model to
use.
I would use a neural network but I'm a bit biased.
Steve
--
EasyNN-plus. Build Neural Networks. http://www.easynn.com
SwingNN. Forecast the Future. http://www.swingnn.com
Neural Planner Software Ltd http://www.NPSL1.com |
|
|
| Back to top |
|
| ... |
| Posted: Thu Jul 03, 2008 3:18 am |
|
|
|
Guest
|
On Jul 2, 4:32 pm, sailor <BJCorne... at (no spam) gmail.com> wrote:
Quote: All,
I would like to predict the sales performance (# of articles sold, #
of articles required) based on the sales data of each store (# of
articles distributed, # of articles sold, regular product, and so on),
but it seems like I don't find my own way in the jungle of statistics.
What would be the best model to use? Should this just be multivariate
regression? I have seen so many terms like 'time series analysis', '
regression analysis' 'neural network', and am not sure what model to
use.
In case you cannot suggest any model, could you tell me how I should
decide what model to use?
thanks in advance,
Cornelis
Dear Cornelius,
What you want to do is to integrate into one equation ...regression
structure ( and the appropriate lags ) AND ARIMA structure reflecting
omitted stochastic series AND intervention detection reflecting
omitted deterministic input series.
The answer is called ROBUST TRANSFER FUNCTIONS .
Regards
Dave Reilly
Automatic Forecasting Systems
http://www.autobox.com
215-675-0652 |
|
|
| Back to top |
|
| |
|
Page 1 of 1
All times are GMT - 5 Hours
The time now is Tue Dec 02, 2008 2:28 am
|
|