Among the rules that are generated after running the association rules algorithm, the ones that have no obvious explanation are the trickiest to use. Note that a rule can only be useful if it can help us discover and understand a new pattern that is expected to eventually lead toward a certain course of action. If that is not the case, and we cannot explain why event X led to event Y, then it is an inexplicable rule, because it's just a mathematical formula that ends up exploring the pointless relationship between two events that are unrelated and independent.
The following are examples of inexplicable rules:
- People who wear red shirts tend to score better in exams.
- Green bicycles are more likely to be stolen.
- People who buy pickles end up buying diapers as well.