IMPLEMENTASI SISTEM NEURO-FUZZY UNTUK PREDIKSI PRODUKSI AIR MINUM DI PDAM SURABAYA

In this research built the application of software informing the prediction of water drinking production in PDAM Surabaya. It is using the neuro fuzzy system with the structure of Modified Adaptive Neuro-Fuzzy Inference System (Mod ANFIS). The learning process of Mod ANFIS uses the hibrid theory algorithm running in two steps. the first step is the thrive step Of running the mechanism of fuzzy inference based on network architecture and the method of Least Squares Estimation (LSE) and the retrieve step using error backpropagation algorithm modified From the research shows that the System of neuro fuzzy structure Mod ANFIS able to recognize the learning error is about 1,98 % and the average error of the prediction result is about 2,245 %