Association MBA - interpretting results

ra2003
ra2003 New Altair Community Member
edited November 5 in Altair RapidMiner
Hello everyone!

I'm doing a market basket analysis. The goal is to figure out whether to stop selling a group product "x", there will be drop in sales of other products.

The group of product "X" is present in 32% of the baskets. But the lift is low (0.7) with correlated products. The same result occurs when working with greater or lesser granularity of grouping.

Can someone help me?

1- The X group contributes to the sale of other groups?

2- For example, can we say that the "H" group will have about 30%(confidence) drop in sales if we remove "x"?

Premisses conclusion suport confidence lift conviction
A x 0,02 0,16 0,38 0,69
B x 0,01 0,16 0,38 0,69
C x 0,02 0,18 0,43 0,71
D x 0,01 0,19 0,45 0,72
E x 0,03 0,20 0,47 0,72
F x 0,02 0,21 0,51 0,74
G x 0,03 0,27 0,63 0,79
H x 0,02 0,32 0,76 0,85
I x 0,02 0,33 0,79 0,87