search

IBMDecisionOptimization / docplex-examples

These samples demonstrate how to use the DOcplex library to model and solve optimization problems.
https://ibmdecisionoptimization.github.io/
Apache License 2.0
396 stars 230 forks source link

Issue on the output of marketing_campaign.ipynb #27

Closed suresh-abeyweera closed 3 years ago

suresh-abeyweera commented 4 years ago

'data' shows all the 2 confidence scores for each customers.If we thoroughly investigate we can notice no customers has higher probability of selecting for Car Loans.But if we analyze the model output we can see some of the Customers are offered CAR Loan,,Practically can this be done? And if we go through the objective function "mdl.sum( channelVars[idx,p,idx2] c.factor productValue[p]* o.Confidence1
for p in productsR for idx,o in offers[offers['Product1'] == products[p]].iterrows()
for idx2, c in channels.iterrows())"

that each decission variable should be multiplied each confidance score,How this can be applied since for Car Loans no probability is given..

vlkong commented 4 years ago

Hi,

I think the person who initially wrote this example overlook the data at hands, and forgot to factor in the confidence for Car Loans (probably because he saw that there was no Car Loan in data !)

This is clearly some missing parts (bug?). However, as there is no Car Loan probability anyway, the output will not change: if all Car Loan probablity are set to 0 (zero), the result will be the same - all customers will be proposed some Car Loans using channel "newsletter".

Thanks for the report, I will follow up with dev team to have this fixed.