Mapping and Prediction Analysis of Rice Fertilizer Use in Paron District, Ngawi Regency using K-Means and Fuzzy Sugeno Methods

Abstract

Clustering and prediction methods are effective tools in overcoming this problem. In this research, the K-Means method was used to map fertilizer needs based on regional characteristics, such as soil type, planting patterns and land productivity levels. This method is able to group regions that have similar characteristics, making it easier to determine optimal fertilizer allocation. Next, the Fuzzy Sugeno method is applied to predict fertilizer needs in the next planting season based on historical data and the results of the grouping that has been carried out. By combining these two methods, it is hoped that the results of this research can provide accurate recommendations for the effective and efficient use of fertilizer. Mapping and predictive analysis on the use of rice fertilizers in Paron District, Ngawi Regency in 12 villages were carried out by referring to the K-Means and Fuzzy Sugeno methods. Mapping with the K-Means Method is done by determining the centroid and clustering, while the prediction mapping using the Fuzzy Sugeno Method is done by determining the variables and fuzzy sets as well as the condition calculation rules. The results obtained are that 2 villages are in the category of low fertilizer use, 3 villages are in the category of medium fertilizer use and 7 villages are in the category of high fertilizer use. Based on the prediction calculation using the Fuzzy Sugeno Method, it was found that the level of error in the comparison of data was 5.36%, meaning that from 100% of the error rate difference, the truth value in calculating the prediction of the use of urea fertilizer using the fuzzy Sugeno method was 94.64%.