Penerapan Metode K-Means Clustering Dalam Pengelompokan Kualitas Tamanan Jeruk Manis

Abstract

Sweet orange is characterized by leaf stalks that have wings and white flowers. Sweet orange plants have stems that can reach a height of 6 m, have many branches, round leaf crowns, and bear fruit once a year. Sweet orange fruit has a round or almost round shape, is large, has strong stems, has a fruit skin that is green to shiny yellow. Sweet oranges contain pectin enzymes which function to reduce LDL (Low Density Lipoprotein) or bad cholesterol, reduce blockage of arteries and reduce the risk of heart disease. Sweet orange fruit also contains flavonoids which can increase the effectiveness of vitamin C and strengthen the walls of blood vessels. One sweet orange fruit contains 16 g of carbohydrates which contain 70 calories and fiber which is equivalent to 12% needed by the body. The lack of sweet orange production is due to virus or pest attacks, some planting centers have experienced a decrease in production. The impact that occurs is a decrease in sweet orange production at a time when there is a lot of demand for sweet oranges by companies and the community. The sweet orange data processed in this study came from the Department of Agriculture and farmer groups in Gunuang Omeh District, 50 Kota District. Furthermore, the data is processed using the Rapid Miner software. Aims to determine the results of the prediction of the quality of the sweet orange planting experiment. The method used to solve this problem is the K-Means clustering algorithm. The K-Means clustering algorithm is a method that attempts to classify existing data into several groups, where data in one group have the same characteristics as each other and have different characteristics from data in other groups. From testing this method, three groups were found in the cultivation of sweet oranges, namely 'good' (can be used as a benchmark for planting sweet oranges so that they can produce good fruit when harvested), 'moderate' (still in the evaluation stage so that it can be good but still usable). for planting trials even though the results are not optimal) and 'not good' (can be used as a reference in order to evaluate the planting structure of sweet oranges in order to anticipate losses during harvest). From the results of the analysis using the K-Means clustering algorithm with the use of the Rapidminer Studio application, three clusters were found, namely C0 with poor results influenced by the age of the seedlings, C1 with moderate results with the influence of spacing and C2 with good results which were influenced by the use of pesticides, the use of organic fertilizers and hoarding.