Topic: Apriori Algorithm& Association Rule Mining

This week we discussed Association Rule Mining, which is initially used for Market Basket Analysis. Market basked Analysis is used to find out how items that customers buy are related to each other, and how frequently they are bought together. All data using the Association Rule Mining are categorized, and it is a poor option for numeric data.

Association Rule has multiple types:

• Actionable Rules
• Trivial Rules
• Inexplicable Rules

We also learned key concepts such as support and confidence.

Support is the rate of frequently bought items or the combination of items bought frequently.

Confidence is the conditional probability of occurrence of a consequent, then providing the occurrence of an antecedent.

The Apriori Algorithm has a two-step approach:

• Frequent Itemset Generation
• Rule Generation

The Steps are:

1. Set minimum support and confidence and calculate the frequency table for the itemset
2. Generate K-itemset from the previous itemset and count each itemset support
3. Repeat for K-Itemset
4. Repeat calculation k-itemset support value
5. Check whether there are more possible itemsets
6. Create Association Rules