Association Rule Mining for Market Basket Analysis: A Comparative Study of the Apriori and FP-Growth Algorithms

Abstract

Association rule mining is a fundamental data-mining task that uncoversinteresting relationships among items in large transactional databases, withmarket basket analysis being its most familiar application. The discovery offrequent itemsets, however, is computationally demanding, and the choice of algorithm strongly affects efficiency. This paper presents a comparative study of the two most widely used algorithms for association rule mining, namely Apriori and FP-Growth, applied to transactional data for market basket analysis. The algorithms were evaluated across a range of minimum-support thresholds in terms of execution time, memory usage, and the number of frequent itemsets and rules discovered. The results show that, while both algorithms generate the same rules for a given support and confidence, FP-Growth is substantially faster and more scalable, particularly at low support thresholds where the number of candidate itemsets for Apriori grows rapidly. Apriori remains simpler to implement and understand. The study confirms the efficiency advantage of the pattern-growth approach and offers practical guidance on algorithm selection for association rule mining. KEYWORDS: Association Rule Mining, Market Basket Analysis, Apriori, FP-Growth, Frequent Itemsets, Data Mining, Support and Confidence.