Neural learning: price prediction for non-agricultural commodities using back propagation network

Neural Network is relatively superlative in predicting economic data. The concept for econometric research contracts with predicting the price variation of non-agricultural commodities. With a focus on gold, silver, aluminium, lead, zinc, natural gas, crude oil the systematic learning for finding the price growth is the aim of this research. The methodology implemented deals with neural network back propagation for training and testing. The input data are learned with 0.2 learning rate and trained until minimization of error. The price of describing commodities from 2007 to 2016 year and it is predicted that within this period the importing commodities get profit up to 77% with combination of gold, silver, aluminium and crude oil are sold as equally. Systematically, the goods are learned easily in the proposed methodology and the error rate is minimized at the lowest.