DATA Step, Macro, Functions and more

creating escalation for bins with sparse data

Reply
Contributor
Posts: 60

creating escalation for bins with sparse data

I am calculating propensity scores for groups of customers. My by groups are :date,  LifeStgDesc, CNXDesc, network, Bundle this gives me 1236 groups adn i am calulating propensity to churn : event/ _Freq_, So waht i want to do is when i have groups with frq< 100 to drop the bundle and replace the propensity of that group with the propensity of the group that has the same date  LifeStgDesc CNXDesc network and max freq (i am not sure yet about the exact escalation i want to use but this is an example).

 

Anyone has any thoughts on this?

 

Thanks,

Myrto

Ask a Question
Discussion stats
  • 0 replies
  • 170 views
  • 0 likes
  • 1 in conversation