<![CDATA[Market Basket Analysis is widely used to identify consumer purchasing patterns from transaction data and support data-driven business decisions. This study aims to analyze product purchase patterns using the FP-Growth algorithm on the Online Retail Dataset obtained from the UCI Machine Learning Repository. The research employed a data preprocessing stage, including removing missing values, duplicate records, cancelled transactions, and invalid transaction values, followed by transaction transformation and one-hot encoding. The FP-Growth algorithm was implemented with a minimum support of 1%, while association rules were generated using a minimum confidence of 50%. The results produced 973 frequent itemsets, indicating that the algorithm effectively identified frequently purchased product combinations. The strongest association rule was found between REGENCY TEA PLATE PINK and REGENCY TEA PLATE GREEN, with a support value of 0.0109, a confidence value of 90.18%, and a lift value of 61.90. These findings demonstrate that FP-Growth is effective in discovering purchasing patterns and can support cross-selling, product bundling, shelf arrangement, and inventory management strategies in the retail sector.]]>
Copyrights © 2026
