Implementasi Data Mining Transaksi Penjualan Menggunakan Algoritma Clustering dengan Metode K-Means

Abstract

The large number of products sold by the Bill Lights Store resulted in a stockpile of several product items due to the large supply of products that were less attractive to customers, resulting in many unsold and under-sold products. Bill Lights struggles with inventory levels of sold and unsold products, as well as shortages and overstocks. Bill Lights stores should rank each product so that they know which products are in the most demand. The purpose of this research is to solve the problem of using inventory information by grouping inventory products based on product characteristics using data mining techniques. The technique used is the K-Means algorithm method. K-Means algorithm clustering method and RapidMiner software processing. The data mining process starts with data processing (selection, cleaning, transformation, data mining and interpretation/evaluation). So if we start with a dataset of 160 products, we get cluster 0 with 88 products classified as sold, cluster 1 with 26 products classified as unsold, and cluster 2 with 46 fewer products classified as sold. The result of using the K-Means method is grouped into three clusters. To enable Bill Lights Store to implement sales and growth strategies based on products that are selling well.

Keywords

• clustering
• k-means
• inventory data