Penerapan Teknik Neural Network dalam memprediksi Perkembangan Impor Kelompok Industri Tekstil dengan Metode Backpropagation

Abstract

The aim of this research is to analyze the development of the textile industry group in Indonesia using Artificial Intelligence. The analysis is conducted through a predictive model that will be used to predict the import development of the textile industry group. The dataset is sourced from the Indonesian Central Bureau of Statistics through the website https://www.bps.go.id/. The technique used is neural network with backpropagation method, and the analysis is conducted using Matlab. Backpropagation is a training method that has a target to be sought. This method is also a multilayer method, which has input, hidden, and output layers. The research process consists of two stages, namely the training stage and the testing stage. Out of several architecture models tested (3-10-1, 3-25-1, 3-50-1, 3-80-1, and 3-100-1), the best architecture model obtained is 3-100-1 with an MSE of 0.000999996 and an accuracy value of 100 percent.