Association rule learning, AKA associations learning. It is used to find interesting relationships or pattern between variables in large datasets.
Types of associations learning
- Market-basket analysis
- Find combinations of items that occur typically together
- Sequential analysis
- Find frequent sequences in data
Market-basket analysis
Market-basket analysis is a data mining technology used to discover association rules between items that customers frequently buy together. The goal is to find what items are commonly purchased together.
Example
From a supermarket data set. If many people buy bread and butter, and often also buy jam, the rule might be: {bread, butter} -> {jam}
This helps stores optimise promotions, store layouts, or recommendations.
Sequential analysis
AKA sequential pattern mining, is a technique used to find patterns in data where the order of events matters - it identifies sequence of actions or purchases over time.
Example
In online shopping or browsing behaviour, a user first visits:
homepage -> product page -> adds to cart -> makes a purchase
A common sequence could be A -> B -> C.