Clustering Potential Area of Fusarium Oxysporum As A Disease of Garlic

Abstract

The decrease of garlic production can be caused by several factors, one of the main factors is a disease. The main disease of garlic is Fusarium Oxysporm (FO) or known as the rotten base. The decrease in garlic production can be reduced by identifying which areas are likely to be attacked by FO. This study aims to determine clusters of the areas in Central Java Province as the largest producer of garlic in Indonesia which is potentially affected by FO so that this disease can be prevented. This study uses the weather data of 2017 including temperature, humidity, and wind velocity for clustering the potential areas attacked by FO. The algorithm used is K-Means clustering. The best clustering results are obtained at number of clusters K=10, it was obtained from several experiments to determine the number of K. The most potential month for FO breeding is October. The results show that 37 % of areas in October 2017 fall into the minimum category, 63 % FO the area belong to the medium attack category and no one is included in the maximum category due to the fact that daily weather conditions do not fit into that category. October became a potential period because that period was the period with the highest number of regions included in the medium category.