on 09-03-2019 6:07 AM
What is it : I have a business scenario where we create around 2000 sale orders a day . I am trying to develop a program/module which predicts the customer orders , analyzing the frequency of orders a customer has made in the past.
For example, a customer places an order for product "P001" every week 1000 nos, and "P002" bimonthly 500 Nos , it should predict what this customer might order the coming week.
What's the benefit ? : We have a tele-sales team who calls the customer and take their orders. If they already have this information with them , it is easy for them to get the orders . Also it helps to give offers and promotions.
What I am looking for : BRAINS
What I have : VBAK ,VBAP.
Sreehari
Hey,
From what I can sense the vibe of the question and your comment to Lakshmipathi's answer, you need to implement an association rule algorithm and some sort of history based analysis like LSTM.
As the actual ABAP implementation for doing so goes far beyond this scope I can only wish you good luck and hope that you will share your code in a blog or a Git repository for us all to benefit from it.
I would add to that and say that you can do it in ABAP core or you can export it to R, Python or other environment that has these packages already implemented do the processing there and return it back to your desired destination.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
This is exactly what I was looking for. My intention is not deferentially to trouble ABAP for this , but to get an idea about the mathematical model I can possibly use for such scenarios. I will give a try in HANA itself ( with PAL ) for this.
My expectation is to understand how I can manually do this prediction , before implementing it .
Sreehari
Why don't you explore Sales Forecast Planning functionality?
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
90 | |
10 | |
10 | |
10 | |
7 | |
7 | |
6 | |
5 | |
4 | |
3 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.