Analisis Penjualan dan Pelanggan Toko Swalayan di Kediri Menggunakan Market Basket Analysis dan Machine Learning

Abstract

Inventory management and sales analysis in convenience stores 
are often handled manually, making it difficult for store owners to 
understand customer purchasing patterns and make effective 
operational decisions. This study analyzes transaction and 
inventory data from a convenience store in Kediri by combining 
Market Basket Analysis and machine learning to produce more 
measurable, actionable insights. 
The methods include FP-Growth to discover product associations, 
Decision Tree to predict inventory status (restock/overstock), 
K-Means to segment products, and Random Forest Regression to 
forecast monthly profit. The results show that consistent 
purchasing patterns can support bundling and product 
arrangement recommendations, while the classification, 
clustering, and regression models help improve stock monitoring, 
wholesale strategy, and financial planning. All outputs are 
implemented in an interactive dashboard to support practical use 
by the store owner.