Clustering of Tidal Swamp Soil Based on NPK Macronutrients using the FCM Method

Tidal swamp resistance is divided into several types, namely type A, B, C and type D. This is divided based on the type of seawater overflow, with the presence of several types, each type has different characteristics, this study aims to enrich the soil, by clustering tidal swamp land using the variables sodium, phosphorus and Potassium (NPK). NPK is a macro nutrient that is needed by plants to grow and develop plants planted on the soil, considering that tidal swamp land is not fertile soil. This research was conducted in the village of Mulia Sari, Tanjung Lago District, Banyuasin Regency with type A. In solving this soil clustering problem, the researcher chose the FCM (Fuzzy C-Miens Clustering) method, because this method is included in the unsupervised learning technique. This method is used to find patterns in a tidal soil dataset. FCM is a popular fuzzy algorithm and can achieve a convergent cluster center. The initial step taken in this study was to measure the NPK macro elements contained in the soil dataset, by taking soil samples to be examined in the soil laboratory, determining the number of clusters, the study determined 3 (three) clusters based on the NPK variable. Then carry out the data processing using the FCM method. From the data processing, the sodium cluster was found in the 15th iteration, while the phosphorus cluster was found in the 17th iteration and Potassium was found in the 16th iteration.