<![CDATA[Indonesia is the fourth most populous country in the world. With a large population, Indonesia is not immune to population problems. This happens because the rate of population growth is not accompanied by the provision of clothing, food, and shelter. In other words, the amount of population growth is not balanced with the availability of natural resources, services, and existing facilities. Therefore, predicting population growth is expected to help the government to overcome population problems. This paper will be using Feedforward Neural Network trained by Particle Swarm Optimization (PSO). PSO algorithm is considered to be able to overcome the weaknesses of the Backpropagation algorithm in training networks. In this study, the predicted error rate is calculated using Mean Average Percentage Error (MAPE). The smallest MAPE results obtained were 0,1599% using 6 input neurons, 4 hidden neurons, 1 output neurons in the network architecture, and the dataset used is the population of Malang City from January 2009 to June 2019. The MAPE results showed that PSO is able to train Feedforward Neural Network to predict the population growth of Malang City.]]>
Copyrights © 2019
