From the course: Machine Learning with Python: Association Rules
Join today to access over 25,500 courses taught by industry experts.
Evaluating association rules - Python Tutorial
From the course: Machine Learning with Python: Association Rules
Evaluating association rules
- [Instructor] Association rules are very powerful. They can reveal interest in and previously hidden patterns within very large datasets. However, not all rules are created equal. For the most part, association rules can be classified as either trivial, inexplicable, or actionable. Most of the rules we generate will be trivial or inexplicable. Identifying which rules are actionable is not a trivial task, no pun intended. It often requires considerable time and effort. With the awareness that one person's trash is another person's treasure, we must decide what criteria to use when evaluating the association rules we create. These criteria will vary from person to person. Luckily, there are several objective data driven metrics that provide us with some guidance as we go through this process. We saw one of them in a previous video. It was called support. Recall that the support of a rule is a fraction of transactions…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.