PREDIKSI JUMLAH PRODUKSI GENTENG DI KEBUMEN BERDASARKAN FUZZY INFERENCE SYSTEM MAMDANI

Abstract

Tile is one product that is in demand by many people. This is a trigger for producers to improve their management. One of the efforts made is to predict the production that can be done to get the optimal amount obtained, so as to get a big profit. In this study, to predict the amount of tile production, computerized calculations were carried out using the Mamdani fuzzy logic method. This method uses the concept of monotonous rules, which are then re-selected by experts in tile production by choosing rules that match the actual situation in the agency. This study proves that predictions made with monotonous rules, then re-selection by experts are able to handle the process of predicting tile production. The results of the comparison of predictions with actual production have an error percentage of 28.45%, with a truth of 71.55% (based on the calculation of the Average Forecasting Error Rate (AFER)). Therefore, when implemented in the Fuzzy Inference System, Mamdani can produce optimal tile production predictions.