Re: Creating RFM variables

Tech-Archive recommends: Speed Up your PC by fixing your registry

Those types of calculations are best fit for an OLAP cube. I would create a
cube with the appropriate aggregations and then create mining models on the
cube. If becomes too cumbersome to create the appropriate model based on
the cube due to the UI limitations, you can always create the correct
queries from the cube that you need for the models and put the results in a
table and mine the resultant table.

--

-Jamie MacLennan
SQL Server Data Mining
This posting is provided "AS IS" with no warranties, and confers no rights.
PLEASE POST NEW QUESTIONS AT
http://forums.microsoft.com/msdn/showforum.aspx?forumid=81&siteid=1
<p.paauwe@xxxxxx> wrote in message
news:28efc60a-8a54-4c34-aa16-f57e562125df@xxxxxxxxxxxxxxxxxxxxxxx
Hello,

Momentarily I'm trying to create recency, frequency, and monetary (RFM)
variables from a tranactional database using a named query in the data
source view. These RFM variables will subsequently be used to create
states as input for the sequence clustering algorithm. As the sequence
clustering algorithm can only take one variable as a sequence attribute,
the RFM variables will need to be combined into one state variable.

This proves to be quite difficult for me. The recency variable should
indicate how many periods it has been since a customer made a purchase.
The frequency variable needs to indicate the total number of purchases a
customer made. So, this value has to be incremented with one at every
purchase ocassion. And the monetary variable has to indicate the total
amount spend. So, the current purchase amount has to be added together
with every past purchase amount.

After these variables are created I want to discretize the values and put
them into one state indicator variable.

All help, tips and hits are very much appreciated.

Greetz and thanks in advance,

Peter


.