Practical Predictive Analytics

Finding the association rules

As shown earlier, the association rules are run using the apriori() function. The apriori() function has several filtering parameters that are used to control the number of rules that are produced. In our example, we will specify the minimum support and confidence threshold that a rule needs to pass in order to be considered.

The number that you pass to apriori depends upon how you want to look at the rules. It can be an initial screening, or it can be a deeper dive, after you have performed several passes. But generally, if we want many rules we can decrease the support and confidence parameters. If we want to focus on items that appear frequently, we raise the support threshold. If we want to concentrate on ...