Penerapan Data Mining Menggunakan Algoritma Fp.Growth Untuk Pengenalan Pola Pembelian Produk Grosir

Abstract

In the retail business that is growing in Indonesia, the most crucial thing is determining the amount of inventory. The inventory of goods is a determining factor in the success of retail companies. So far, inventory checks have only been carried out manually. So this becomes inefficient when consumers need goods in large quantities and the stock does not meet their needs. The purpose of this study is to apply a data mining method by analyzing consumer purchasing patterns based on sales transaction data. This study uses a mixed-methods design (mix-method) of quantitative and qualitative methods. The material used in this study is product sales transaction data from 2021, which amounted to 1,580 transactions. After cleaning, 594 data points were obtained that were ready to be tested. The tools used for testing are rapidminers. In this study, the author uses the CRISP-DM model, which is used for data mining processing. From the calculation process of the fp-growth algorithm with a minimum support parameter of 10% and a minimum confidence of 50%, the association rules formed are 6 rules. Of all the rules, one was formed with the highest confidence value of 87%, the support value of 12%, and all the rules had a lift ratio value of > 1, which means that the rule is valid and can be used properly. These results can be used to determine the retail company's sales strategy