Clustering of Featured Vegetables Using the K-Means Algorithm

Abstract

Horticulture, especially vegetables, has great potential to be developed because it becomes a source of income for the community and small farmers in each region because Indonesia is called an agrarian country with most of them working in agriculture. Mandailing Natal Regency is the district with the largest area in North Sumatra province, but Mandailing Natal has not been able to outperform vegetable crop production in North Sumatra. Data mining methods can find interesting and invisible patterns in data sets. One of the methods is the K-Means clustering algorithm which groups data into clusters based on the similarity of data characteristics. In this study, vegetable data was clustered which aims to determine the potential commodities in each area in Mandailing Natal Regency, plants that have potential in the area will be maintained and their production increased, while vegetable crops whose production is still low will be a priority to increase their production. The research method used in this study was to collect vegetable data from the Badan Pusat Statistik in the form of data on harvested area, production, plant area, and new planting area. In addition, data collection was carried out by conducting theoretical studies in journals. The results of clustering superior vegetables using the K-Means Algorithm are in the form of potential grouping into 3 clusters, namely low, medium, and high clusters and the output is a web-based system in its application. The results of the clustering analysis were obtained with each total data of 69 data, namely big chili with C1 81%, C2 16% and C3 3%. Cayenne C1 29%, C2 48% and C3 23%. Long Beans C1 26%, C2 38% and C3 36%. Kale C1 39%, C2 36% and C3 25%. Eggplant C1 43%, C2 29% and C3 28%. Tomato C1 41%, C2 58% and C3 1%.