Analysis and Prediction of Rural Residents' Living Consumption Growth in Sichuan Province Based on Markov Prediction and ARMR Model

I select 32 samples concerning per capita living consumption of rural residents in Sichuan Province during the period 1978-2009. First, using Markov prediction method, the growth rate of living consumption level in the future is predicted to largely range from 10% to 20%. Then, in order to improve the prediction accuracy, time variable t is added into the traditional ARMA model for modeling and prediction. The prediction results show that the average relative error rate is 1.56%, and the absolute value of relative error during the period 2006-2009 is less than 0.5%. Finally, I compare the prediction results during the period 2010-2012 by Markov prediction method and ARMA model, respectively, indicating that the two are consistent in terms of growth rate of living consumption, and the prediction results are reliable. The results show that under the similar policies, rural residents' consumer demand in Sichuan Province will continue to grow in the short term, so it is necessary to further expand the consumer market.