中国城镇居民消费支出的多元非线性回归模型研究

Using Multiple Nonlinear Regression Model to Study Consumption Expenditure of Urban Residents in China

以1993-2007年数据为根据,采用逐步回归方法,建立中国城镇居民消费支出的多元非线性回归模型.结果表明:影响居民消费支出的主要因素有收入、消费意愿、居住面积、商品零售价格.消费支出随着收入、消费意愿、商品零售价格的提高而提高,随着人均居住面积的增加而先增后减.而且多元非线性回归模型比线性回归模型更能准确描述客观实际结果.

Based on the data of the 1993-2007 annual data,and using stepwise regression method,we establish the multi-consumer spending,non-linear regression model.The results show that：the residents are a major factor in consumer spending,income,consumption will,living space,commodity retail prices.Consumer spending as incomes,consumption will increase the retail price increase,with an increase in per capita living space increased after the first cut.And multi-nonlinear regression model is more accurate than linear regression model for describing the results of objective reality.