3

I want to compare attribution results made by Markov Chain and Shapley Value. My data set contains 2 columns (path and nb_of_conversions):

path=c('Paid Search > Direct > Paid Search','Organic Search > Display (impression) > Display (impression) > Display (impression) > Direct','Referral > Direct','Direct > Organic Search','Paid Search > Display (impression) > Display (impression)','Paid Search > Direct > Direct > Direct','Display > Display','Direct > Direct > Paid Search')
conversions=c(77,125,68,50,32,29,43,84)
data<-data.frame(path,conversions)

For Markov Chain there is a good library 'ChannelAttribution', so I have no problems with it. But I'm struggling with Shapley value.

How can I built my code to attribute conversions to each tochpoint using Shapley value princips?

user3825422
  • 31
  • 4
  • To the extent that this asks for code (or how to code something, or what functions to call) it's probably [off topic](http://stats.stackexchange.com/help/on-topic) here. If you edit to focus your question on the underlying issue (the problem you're solving and the particular methodologies that might solve it) it might be on topic. – Glen_b May 02 '17 at 01:01
  • Check out the blog below, it is a clear example how to implement it. http://datafeedtoolbox.com/attribution-theory-the-two-best-models-for-algorithmic-marketing-attribution-implemented-in-apache-spark-and-r/ – Joep van der Plas May 01 '18 at 12:31

0 Answers0