四川省农村居民生活消费增长分析及预测——基于Markov预测和ARMA模型

选取1978～2009年四川农村居民人均生活消费值的32个样本,首先,通过Markov预测法预测未来生活消费水平的增长速度以10%～20%的概率较大;然后,为提高预测精度,在传统ARMA模型中加入时间变量t进行建模并预测,预测结果表明平均相对误差率为1.56%,其中2006～2009年的相对误差的绝对值均小于0.5%;最后,将Markov预测和ARMA模型对2010～2012年的预测结果对比,发现两者在生活消费增长幅度上吻合,预测结果可靠。结果表明,在与目前相似的政策力度下,短期内四川省农村居民消费需求将持续增长,需进一步扩大消费市场。

Analysis and Forecast of the Living Expenditure Increase of Rural Residents in Sichuan Province Based on Markov Forecast and ARMA Model

Based on the 32 samples of per capita living expenditure of rural residents in Sichuan from 1978 to 2009,it is forecasted by the Markov method that the level of living expenditure will probably increase at a rate of 10%-20%.In order to improve the precision of forecast values,the time variable of t was added to traditional ARMA model,and the forecast results showed that mean relative error rate was 1.56%,and the absolute value was less than 0.5% from 2006 to 2009.By comparing the forecast results from 2010 to 2012 by Markov forecast method and ARMA model respectively,it was concluded that their increase ranges of living expenditure were basically consistent,and the forecast values were credible.The expenditure of rural residents will continue to increase under similar policies by the forecast results and it is necessary to further expand the consumer market.