Posted: 16 May 2018 6:15 EDT Last activity: 29 May 2018 7:16 EDT
I cannot find enough material on the champion challenger topic to understand its full working. The only material I could find is in the practice section in the Pega Decision Consultant. Can anyone help me in finding some relevant material on the champion challenger topic in pdn.
The champion challenger component basically just generates a random number and selects between its inputs based on that
It is typically used to test a model, strategy or treatment on a small, random, set of your population. Then after monitoring the results you change the percentages or redesign the propositions, strategies or models if they're not giving you the performance/success rate that you're after.
If you want to do something more 'automated' then you would use something like a priority component typically. Say as an example we have multiple cross sell offers for a customer, and adaptive models that calculate the likelihood of a response, a priority component can the output the top (n) offers with highest predicted response. Ofcourse you can throw other weightings into the mix to calculate priority, and also exclude offers (decision alternatives) based on hard exclusion rules.