Penerapan Algoritma Apriori dan FP-Growth Untuk Market Basket Analisis Pada Data Transaksi NonPromo

This research aims to find association rules based on the transactions of Aksesmu members on non-promo items. The method in this study uses Association rules using the a priori algorithm and FP-Growth to obtain Frequent Itemsets. The data analysis phase is carried out starting with Exploratory Data Analysis, Pre-Processing Data, Transformation Data, and Data Mining, to evaluate the results of the formed association rules. Researchers conducted 4 experiments with a minimum support of 0.02 and a minimum confidence of 0.25 on a priori and FP-Growth was the best by producing 52 frequent itemsets and 17 association rules. With a dataset of 379,635, a priori is faster in processing frequent itemsets with a time of 1.10 seconds while FP-Growth is with 1.86 seconds. Apriori and FP-Growth produce the same frequent itemset, namely the highest category is obtained by SKT with a support of 0.32 and SKM with a support of 0.26, but the best association rules are produced by the Extruded & Pellet and Sweetened Condensed Milk categories with a confidence of 0.47, which if items in the Extruded & Pellet category are purchased together with Sweetened Condensed Milk category items have a success rate of 47%.