🎉 🎉 RAPIDMINER 9.10 IS OUT!!! 🎉🎉

Download the latest version helping analytics teams accelerate time-to-value for streaming and IIOT use cases.

CLICK HERE TO DOWNLOAD

i need help in decision tree

WaadWaad Member Posts: 7 Learner I
edited July 13 in Help
Hi everyone,
i had this problem i should do with decision tree and i did this process attached (refundDT1) this was the result


and i got this :
"for which channel there are canceled orders and the associated early detection of possible cancellation sources in the channels and process steps" You create a decision tree with only one descriptive variable (variant index) and target variable "Channel". Where can you read "which channel has canceled orders"? And where are the sources of cancellation? As with 1), it is wrong to refer to the variant index. The variant index is just a list of the different variants and its level has no relation to the channel or the termination

now i made this new process (DTrefund2.rmp) and i get the same result
so what should i do and why is it using Variant index only?

Thanks :)


Tagged:

Answers

  • jacobcybulskijacobcybulski Member, University Professor Posts: 391   Unicorn
    It seems that none of your attributes, apart from the variant, are differentiating one channel from another. If you look at your data, all other attributes relate to the case, which defines the channel and which has all stages of processing always present (at different dates). So none of those attributes are helpful in decision making. You may need to extract your data from the enterprise system in a different way, perhaps look at the normalised tables to get some meaningful outcome. 
Sign In or Register to comment.