利用神经网络对居民消费价格指数预测的研究——以广州为例

PREDICATION OF THE CONSUMER PRICE INDEX BY BP NEURAL NETWORKFOR GUANGZHOU

居民消费价格指数是经济地理研究的重要依据,但其运算是一个非常复杂的过程。采用归一化数据处理方法,选择神经网络的训练样本,建立基于BP神经网络的居民消费价格指数预测的数学模型。利用该模型对广州居民消费价格指数进行预测,结果表明模型的预测值与实测值的误差仅为0.91%。用神经网络来构建模型后进行预测,是一种新方法的尝试,通过检验也是可行的。用MATTLAB语言编写程序,对数据进行处理后,能预测到未来的发展情况,对相关方面起到指导作用。

Consumer price index is the important basis of economic geography, but operation is a very complicated course. Adopted the data processing method of the normalization, choose the training sample of the neural network, the mathematical model of the consumer price index based on BP nerve network predicts set up. With this model, the average absolute error between the predicted data and experimental data was only 0.91%, indicating that the model has high predicting accuracy and is worth of practical application. After writing the procedure in MATrLAB language and dealing with the data, the model can predict to the future development and play a guiding role in relevant respects.