Pastry Sales Data Clustering with K-Means Clustering Approach for Product Grouping

Abstract

In the business world that is run by many people today, we are required to always develop our business so that it always develops to make a profit. To achieve this, there are several things that can be done, namely by increasing product quality, adding product types, and reducing company operational costs by using company data analysis. Mr Dosi's Pastry Shop is a shop engaged in the sale of packaged dry food. Data on the purchase of MSME packaged pastry stock was not taken into account, which then led to a buildup of packaged food which made sales turnover less effective. Packaged pastries are a type of food that is packaged in packs per kilogram. Efforts are needed to find out the causes of sales that have decreased so that sales targets cannot be achieved. Related to the process of grouping the data, a data mining technique is used to perform clustering. This clustering process can be done using K-Means Clustering. This algorithm performs grouping of data sets into a predetermined cluster which aims to form separate data groups that have similarities. The results of the clustering of product sales levels are then used as a reference in warehouse management. Based on the grouping results, the grouping with 2 clusters is the most optimal grouping result with the smallest Davies-Bouldin Index (DBI) value, namely 0.125.