The Prediction Research on the Transport & Storage of Guangdong in China with the GMDH Algorithm Method

This paper uses GMDH method to establish a prediction model to forecast the output value of transport & storage of Guangdong in China, since the original samples of the output value of transport & storage are less enough to be used with the traditional methods. Compared with traditional linear regression and artificial neural network, the predicted results show that GMDH method is an effective way to predict car ownership with a small sample of history data in the complex economic system.