The Altair Community is migrating to a new platform to provide a better experience for you. In preparation for the migration, the Altair Community is on read-only mode from October 28 - November 6, 2024. Technical support via cases will continue to work as is. For any urgent requests from Students/Faculty members, please submit the form linked here
Real business problem
Learner III
Dear community,
I am new to data mining/RM and would need a little help to start :-[
I have an actual business problem and would like to use DM for this. The business question is this: Which customer should I offer what product at what discount rate. Since I would like to focus on declining accounts I would also need to include the growth rate of the respective accounts. My data set includes account# (id, nominal), product (nominal), average discount rate (numerical), growth rate (numerical). I can also add geographic location but the importance is somewhat lower (it might be that in some locations people get unusually high discounts). I thought I should segment the customers so that I can focus on some customer groups (e.g. medium discount, medium decline, which products have been bought.
So what do you think would be the best way to go about that and why?
Thanks a lot!
Marcus
I am new to data mining/RM and would need a little help to start :-[
I have an actual business problem and would like to use DM for this. The business question is this: Which customer should I offer what product at what discount rate. Since I would like to focus on declining accounts I would also need to include the growth rate of the respective accounts. My data set includes account# (id, nominal), product (nominal), average discount rate (numerical), growth rate (numerical). I can also add geographic location but the importance is somewhat lower (it might be that in some locations people get unusually high discounts). I thought I should segment the customers so that I can focus on some customer groups (e.g. medium discount, medium decline, which products have been bought.
So what do you think would be the best way to go about that and why?
Thanks a lot!
Marcus
0
Answers
the quesion is kind of complex.
The key question is, if you have "correct' example = labeled data. Then you can go for "standard" supervised learning. Often this is not the case and you need to work in an unsupervised manner with Clustering, Cross-Distances or FP-Growth.
I don't know how big your project is, but Rapidminer is also offering consulting and has quite some partners which might help. If you need any information on that, you can contact me via mschmitz at rapidminer.com
Cheers,
Martin
Dortmund, Germany