ANALISIS POLA TRANSAKSI PENJUALAN UNTUK PENATAAN TATA LETAK PRODUK MENGGUNAKAN ALGORITMA FP-GROWTH PADA TOKO RITEL

Abstract

Strategic product layout organization across retail stores stimulates impulse buying and enriches consumer convenience. This study excavates market basket purchase associations utilizing the Frequent Pattern Growth (FP-Growth) algorithm, comparing execution time and memory footprints against the traditional Apriori algorithm. The empirical corpus comprises 1,500 point-of-sale transactions covering 13 FMCG categories across varying minimum support (2%-5%) and confidence thresholds (60%-80%). Computation results reveal FP-Growth superior speed (46x faster at 15.3 ms vs 804.5 ms) and 12-fold lower memory consumption relative to Apriori. Primary association rules detected prominent item pairs including Bath Soap -> Powder Detergent (support 13.93%, confidence 69.67%, lift ratio 4.75) and Bread -> UHT Milk (lift 3.58). The resulting association rules guide store cross-merchandising layouts and promotional bundle strategies.