Fuel Oil Price Forecasting Using Symbiotic Evolutionary Immune Clustering Neural Network

Oil price time series is a nonlinear long-memory series, in this paper, a novel clustering method based on the symbiotic evolutionary and the immune programming algorithm is proposed, which is implemented for the prediction of oil price time series. In the design of the neural network, the number and positions of hidden layer are automatically adjust through symbiotic evolutionary and the immune programming. The weights of output layer are decided by the recursive least squares algorithm. The proposed immune clustering neural model has been implemented for New York harbor residual fuel oil prices, and compared with the traditional RBF neural network method. The test results reveal that the symbiotic evolution immune clustering neural network method possesses far superior forecast precision than the traditional method.