JIMT Vol. 14 No. 1 Juni 2017 (Hal 95 - 106) 95 ISSN : 2450 – 766X APLIKASI MODEL NEURO FUZZY UNTUK PENGONTROL TINGKAT INFLASI DI PROVINSI SULAWESI TENGAH Rivaldi 1 , R. Ratianingsih 2 dan D. Lusiyanti 3 1,2,3 Program Studi Matematika Jurusan Matematika FMIPA Universitas Tadulako Jalan Soekarno-Hatta Km. 09 Tondo, Palu 94118, Indonesia. 1 Rivaldibaharat@gmail.com, 2 ratianingsih@yahoo.com, 3 Desylusiyanti@yahoo.com ABSTRACT Neuro fuzzy models is the merging of two systems, that is Artificial Neural Network (ANN) and fuzzy logic. ANN is a structure that mimics the presence of neurons (nerve) as well as in the human brain. While fuzzy logic (fuzzy logic) is the use of membership functions to determine how large a predicate fulfill a function. The purpose of this study was to develop a model neuro fuzzy or Adaptiveneuron fuzzy inference system (ANFIS) to control the inflation rate in the province of Central Sulawesi. Controlling is done after a prediction of inflation indicators used to predict is the Gross Regional Domestic Product at Current Market Prices (Nominal GDP) and Regional Domestic Product Gross Constant Prices (Real GDP) Central Sulawesi Province First Quarter 2011 - Fourth Quarter 2015. Architecture built in this study consisted of 2 architecture, the architecturebackpropagation and the architecture of neuro fuzzy. The use backpropagation architecture in research aims to predict the Nominal GDP and the Real GDP which is then used as an indicator to predict inflation on ANFIS. Data in this study were divided into two parts, that is training data and testingdata with a composition of 70% versus 30%. In this study Backpropagation and ANFIS method is seen as an algorithm that is able to handle complex issues and complex due capable of adapting to changes and uncertainties that accompany issues. Furthermore, the control of inflation data that does not match the prediction results PMK 66 / PMK.011 / 2012. Average error obtained in the process of prediction forecast data is 3.16% ofNominal GDP, 12.16% forecast on the Real GDP and 3.77% in inflation predicted results obtained while in the process of controlling the average error in controlling the process is at 5.14%. The results showed that the design of the network that was formed has been quite good in predicting and controlling the movement of the inflation rate fluctuated. Keywords : Anfis, Backpropagation, Gross Regional Domestik Produk, Inflation, Neuro Fuzzy. ABSTRAK Model neuro fuzzy adalah penggabungan dua sistem, yaitu Artificial Neural Network (ANN) dan fuzzy logic. ANN adalah suatu struktur yang meniru keberadaan sel-sel neuron (syaraf) sebagaimana dalam otak manusia. Sedangkan fuzzy logic (logika fuzzy) adalah pemakaian fungsi keanggotaan untuk menentukan seberapa besar suatu predikat memenuhi suatu fungsi. Tujuan penelitian ini adalah mengembangkan model neuro fuzzy atau Adaptive neuron fuzzy inference system (ANFIS)untuk mengontrol tingkat inflasi di Provinsi Sulawesi Tengah. Pengontrolan dilakukan setelah dilakukan prediksi terhadap inflasi dengan indikator yang digunakan untuk memprediksi adalah Produk Domestik Regional Bruto Atas Dasar Harga Berlaku (PDRB ADHB) dan Produk