城镇居民消费水平经验模型构建

Empirical Model Construction of Urban Residents′ Level of Consumption

从经济学和统计学视角,本文选取影响城镇居民消费水平的因素,并以相关消费理论和模型为基础,利用主成分分析方法和岭回归分析方法建立城镇居民消费水平经验模型。分析结果表明两种方法各有千秋,主成分分析方法在模型拟合方面更加具有优势,拟合结果更加精确稳定,而岭回归分析方法在模型调试方面更加有效。当然,将两者结合起来运用会使经验模型与理论模型更加贴近,更有效地揭示经济现象。

From the perspective of the economics and statistics,we select the factors of affecting urban residents′ level of consumption.On the base of relative consumption theories and models,we construct the model of urban residents′ level of consumption,using the methods of principal component analysis and ridge regression analysis.As a result,they are both relatively excellent.Principal component analysis,however,is much better to simulate total scale,and makes fitted values more accurate and more stable;on the other hand,ridge regression analysis is capable of debugging models very well.Of course,combining them will make empirical models and theoretical models close,and will more effectively reveal economic phenomena.