Practical Predictive Analytics

Plotting many rules

Plots are especially helpful for scenarios in which there are many rules generated, and you need to filter on specific support and confidence ranges.

Here is a plot that shows two of the three metrics, along the x and y axis, and the third metric (lift, support, or confidence) as shading:

sel <- plot(many_rules, measure = c("support", "confidence"), shading = "lift",  interactive = FALSE)

As the following plot suggests, there is a cluster of rules with high lift (>8), high confidence (> 0.6), but all with low support: