Open AccessForecasting Using Back Propagation with 2-Layers Hidden
Author(s) -
Syaharuddin Syaharuddin,
Dewi Pramita,
Toto Nusantara,
Subanji Subanji
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1845/1/012030
Subject(s) - inflation (cosmology) , inflation rate , matlab , welfare , government (linguistics) , econometrics , economics , order (exchange) , function (biology) , computer science , monetary policy , macroeconomics , finance , physics , market economy , linguistics , philosophy , evolutionary biology , biology , theoretical physics , operating system
A critical indicator of government policy in economics is based on high inflation because it implicates the welfare of the society in a region. Therefore, it is necessary to have a computing system to see the inflation trend or pattern by looking at monthly data each year. This research aims to predict the rate of inflation using the ANN Back Propagation method based on GUI MATLAB with two hidden layers. Input data is the inflation data of NTB province by using monthly data for the last 11 years, the activation function is LOGSIG, and the training function is TRAINRP. The results of training and testing were obtained an average prediction of 0.213 with MSE of 0.0053. These results proved that inflation rate of 0.5% decreased, therefore catagorized on medium inflation, with the highest inflation in April and the lowest one in July. The results showed that the Back Propagation method can be used as an alternative in predicting the rate of inflation in order to support sustainable economic growth so as to improve people’s welfare in the future.