Open AccessEstimations of Indonesian poor people as poverty reduction efforts facing industrial revolution 4.0
Author(s) -
Anjar Wanto,
Jaya Tata Hardinata
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/725/1/012114
Subject(s) - poverty , indonesian , government (linguistics) , backpropagation , indonesian government , artificial neural network , developing country , computer science , economic growth , economics , machine learning , linguistics , philosophy
Indonesia is one of the developing countries that have serious problems with poverty. The still many poor people in Indonesia encourage the government to make and determine the right policies so that the problem of poverty can be overcome and not drag on. Therefore, the authors conducted this study to try to help the government conduct an analysis in predicting the level of development of the poor in Indonesia. The prediction method used is the Bayesian Regulation artificial neural network. This method is a development of the backpropagation method that is often used to predict data. The data used are data on poor people in Indonesia in 2012-2018, which are sourced from the Indonesian Central Bureau of Statistics. Based on this data a network architecture model will be formed and determined using the Bayesian Regulation method, including 10-5-10-2, 10-10-10-2, 10-10-15-2, 10-10-20-2, 10-15-10-2, 10-15-15-2, 10-15-20-2, 10-20-20-2, 10-25-25-2 and 10-30-30-2. From these 10 models after training and testing, the results show that the best architectural model is 10-25-25-2. The accuracy of the architectural models is 94.1% and 61.8% with MSE values of 0,00013571 and 0,00005189. The results of this study are the prediction of the poor for the next 5 years.