PREDIKSI INFLASI INDONESIA DENGAN MODEL ARTIFICIAL NEURAL NETWORK

Diah Wahyuningsih, Idah Zuhroh, - Zainuri

Abstract


This research examines and analyzes the use of Artificial Neural Networks (ANN) asa forecasting tool. Specifically a neural network’s ability to predict future trends ofinflation is tested. Accuracy is compared against a traditional forecasting method,multiple linear regression analysis. Finally, the probability of the model’s forecastbeing  correct  is  calculated  using  conditional  probabilities. While  only  brieflydiscussing  neural  network  theory,  this  research  determines  the  feasibility  andpracticality of using neural networks as a forecasting tool for inflation in Indonesia.This study builds upon the work done by Edward Gately in his book Neural Networksfor Financial Forecasting. This research validates the work of Gately and describesthe  development of  a  neural network  that  achieved an  86  percent probability  ofpredicting an  inflation  rise, while multiple  regression  analysis  is only  to predictinflation that achieved a 16%.  It was concluded that neural networks do have  thecapability to forecast inflation and, if properly trained, we could benefit from the useof this forecasting tool.

Keywords: neural networks,  inflation,  time  series analysis,  forecasting, artificialintelligence


Full Text:

PDF


DOI: http://dx.doi.org/10.21776/ub.jiae.2008.002.02.7

Refbacks

  • There are currently no refbacks.