Intelligent system for predicting the price of natural gas based on non-oil commodities

We present a preliminary investigation into a novel approach to natural gas prediction. Experimental data were extracted from the Energy Information Administration of the US Department of Energy. The datasets were pre-processed and used to build a feed-forward neural network intelligent system for predicting natural gas prices based on gold, silver, soy and copper. The validation of the intelligent system indicated a Regression (R) = 0.79972 when the reserved datasets were tested on the intelligent system. Natural gas prices can be predicted using non-oil commodities as independent variables. With little additional information, the proposed design can be used to construct intelligent decision support systems to support decision making in the government and private sector.