Oil price forecasting based on self-organizing data mining

The fluctuation of oil prices attracts the great attention of the world. However, the prediction of oil prices is very difficult because the oil price system is so complex. In this paper, AR-GMDH algorithm and AC algorithm are adopted to forecast oil prices. The validity and feasibility of self-organizing data mining are manifested by the comparisons of the prediction result with that of conventional statistical methods. The result shows that self-organizing data mining is a precise method to forecast such complex systems.