The Clustering of the Cooperative of the Republic of Indonesia Employees in Bangkalan Using the K-Means Algorithm

Abstract

Cooperatives are business entities that drive the community's economy with a business model owned collectively by their members. The Cooperative of the Republic of Indonesia Employees (KPRI) is the most common type of cooperative among active cooperatives in Bangkalan, with savings and loan operations as its main business activity. This research aims to classify KPRI based on its financial ratios using the K-Means Clustering algorithm. The data collected for this research consists of 47 financial data points from the KPRI, obtained from the Annual General Meeting reports of the cooperatives available at the Cooperative and Micro Business Office of Bangkalan Regency. Data preprocessing used the Correlation Matrix method and was then normalized using MinMax scaling. The K-Means method requires a value of k, which in this research was determined using the elbow method. The optimal value of k found was 3, indicating that three clusters of KPRI were generated. The results show the characteristics of each group based on their financial ratio conditions. The analysis of the average ratio in each group showed that Group 0 has weaknesses in the return on assets ratio and the return on equity. Group 1 has weaknesses in the current ratio and return on equity, while Group 2 has weaknesses in the current ratio and all types of profitability ratios. In the clustering results, Group 2 comprises 51.06% of the total data, meaning that half of the active KPRI in Bangkalan have issues with their liquidity and profitability ratios. Meanwhile, Group 0 contains 36.17% of the members, and Group 1 has 12.76%.