Prediction of Natural Rubber Customs Declaration Price Based on Wavelet Decomposition and GA-BP Neural Network Group

Natural rubber is mainly dependent on import in China, its domestic market price influenced by the Rubber Customs Declaration Price (NRCDP). Considering fluctuating properties of NRCDP, a method NRCDP based Wavelet and optimized Back Propagation (BP) neural network Group using Genetic Algorithm (W-GA-BPG) proposed. First, an dataset established original (CDP) collected Qingdao Customs, which commodity types are selected consistently according to sampling intervals, features deleted if they less affected fluctuation NRCDP. Secondly, decomposed wavelet transform obtain group feature sequences with different scales. Then, BP networks (BPG) (GA) used predict multiple decomposition sub-sequences, respectively. Finally, predicted values obtained through reconstruction. Combined dataset, W-GA-BPG model comparing analyzing experiments evaluating Mean Square Error (MSE) determination coefficient prediction results. The MSE proposed 0.0043 0.9302, respectively, best effect.