Urbanization level forecast based on BP neural network

Urbanization is the general trend and tide of the current world development and also one of the most remarkable social and economical phenomena in the world. Urbanization level becomes an important symbol of the economic strength and modernization level in a region and thus how to improve the local urbanization level has become a priority for economic development. With the increasing urbanization level, a series of adverse effects have been brought about. It is a very important task to make a scientific forecast for the urbanization of a region. Take Zhejiang Province as an example; only considering the present economic development factors, the currently most popular BP network (Back-Propagation Network) in the neural network is adopted to establish the forecast model; and the urbanization level of Zhejiang Province in 2000-2004 is forecasted. For the forecast result, the maximum relative error is 2.51%, the minimum relative error is 0.23% and the mean absolute percent error is 1.05%. Thus, the result indicates that the model has high forecast precision and therefore can be used to forecast the urbanization level of Zhejiang Province in the future.