Help to interpret result from a Lift Chart?

• Sunday, February 03, 2013 7:43 PM

   

   
  

  0

  Hi All,

  I was reading this article from Microsoft site: http://msdn.microsoft.com/en-us/library/ms175428(v=sql.100).aspx

  I need help to interpret the result.  Here is the Mining Legend from the chart below.

  Series, model	Score	Target population	Predict probability
  Targeted mailing all	0.71	47.40%	61.38%
  Targeted mailing under 30	0.85	51.81%	46.62%
  Random guess model		31.00%
  Ideal model for: Targeted mailing all		62.48%
  Ideal model for: Targeted mailing under 30		65.28%

  
  

  It says that "From these results, you can see that, when measured at 30 percent of all cases, the general model (Targeted mailing all) can predict the bike buying behavior of 47.40% of the target population. In other words, if you sent out a targeted mailing to only 30 percent of the customers in your database, you could reach slightly less than half of your target audience. If you used the filtered model, you could reach about 51 percent of your targeted customers." 

  This means that if we use the basic model (no filter) and assume the population is 1,000 and we sent 300 mails (for cases that have a predict probability of 61.38% or higher) out, you will get 142 [300 x 47.4%] responses back.  This part makes sense.

  But what if I sent out 1,000 (100% of overall population), from the line it shows that the target population is 100%.  That means if I sent 1000 mails out I will get 1000 responses?  I know this is not right, but how can we interpret this?

  ---

  BI Analyst

  • Reply
  • Quote