Prediksi Indeks Harga Konsumen (IHK) Kota Ambon Menggunakan Elman Recurrent Neural Network (ERNN)

Abstract

Indeks Harga Konsumen (IHK) is an economic indicator that can provide information on developments and changes in the prices of goods and services that are predominantly consumed by the public within a certain period of time. In this study the method to be used is the Elman Recurrent Neural Network (ERNN). The research data uses Ambon City IHK data from 2016 to 2019. The data used as research objects are: Food, Beverages, Cigarettes and Tobacco, Housing, Water, Electricity, Gas and Fuel, Clothing, Health, Education, Recreation, and Sport, Transportation, Communication and Financial Services as input variables. The results of training with 5 hidden layers at a maximum epoch of 100,000 obtained the smallest MAPE value of 1.1773. Then the results of testing using the parameters in the experiment on the number of hidden layer neurons 20 obtained the smallest MAPE value of 0.461823.