关键词: 消费价格指数, 优化模糊神经网络, 趋向预测

Abstract: The consumer price index is an important basis of consumer price index policy, salary policy, and national economy development strategies. Many researchers have carried out a lot of researches for CPI, and a lot of achievements have been got, however, past predicting method applies the simple neutral network method, and the good predicting results are difficult to be obtained. The paper puts forward a fuzzy rule self adaptive training algorithm according to the existing problems. and the fuzzy rules of different numbers are produced according to the actual problems, and the network structure is confirmed according to fuzzy rule. For the confirmation of the network structure, the test needs not be carried again and again because it can not be formed randomly. Therefore it is responsible. Simulation results show that this method gets the good effect for predicting consumer price index.

Key words: consumer price index, optimized fuzzy neural network, trend forecasting