Implementation of K-Means Clustering Algorithm to Determine the Best-Selling Snack In Purwokerto MSMES

This research was conducted to provide information related to the sale of existing MSME snacks, which products have many buyers and which are not, besides that this research can provide a view for the sale of various snacks whether they are still sold or selling the same snacks but with new innovations so that there are many enthusiasts. To do this, a grouping is needed, therefore the researcher chooses a k-means algorithm which will later be used for the clustering process. The grouping is divided into two, namely best-selling and under-selling products, for research data collected from January to March 2024. Then the data used includes the name of the snack, the number of stocks and the number of sales. After calculating the results of this study, the k-means algorithm performs a calculation of as many as two rounds so as to form two clusters where in cluster one the names of snacks that sell well are grouped such as cimol, noodle skewers, dimsum, egg rolls and white bread. Then in cluster two, the less selling product falls to seblak snacks so that seblak snacks can make new innovations to sell better and can compete with other products. This research succeeded in grouping and providing an overview of products that sell well or not, for future research can be reproduced related to the data used.