PREDIKSI JUMLAH PRODUKSI GULA DENGAN METODE ADAPTIVE NEURO FUZZY INFERENCE SYSTEM

Abstract

Sugar is a staple food consumed by Indonesians, making it essential to optimize sugar production to meet the population's needs. This research will design a prediction system for the amount of sugar production at PTPN XI PG Prajekan using the Adaptive Neuro-Fuzzy Inference System (ANFIS) method. ANFIS is an integrated approach that combines the fuzzy logic system with the artificial neural network system. This research involves several steps: data collection, designing the ANFIS system, training and testing the ANFIS model, calculating accuracy, and analyzing the results. The prediction system for the amount of sugar production is designed to predict the variable y_(t+1) which is the amount of sugar production in the year (t+1) using the input variables x_(1,t) (sugarcane harvested area in year t), x_(2,t) (amount of sugarcane in year t), x_(3,t) (average of yield in year t), and x_(4,t) (number of milling days in year t). The experiments in this study utilized variations in both the types and the quantities of membership functions. The best model obtained in this research is a model with a difference between two sigmoidal membership functions, and a product of two sigmoidal membership functions with a total of 2 membership functions for each input variable. Both models yield identical Mean Absolute Percentage Error (MAPE) values, with 1.79% during training and 4.82% during testing.