Study on Rural Residents Income Growth Based on Quasi-Stepwise Regression Model
To analyze rural residents’ income growth in Heilongjiang province, the “quasi-stepwise regression” variable selection method is proposed. This method is based on the regression estimate of minimum sum of absolute errors. In its application to rural resident revenue forecast, household business revenue model and wage revenue model are established. The total relative error of the fitted model is about 8 %. According to these models, factors affecting rural resident revenue are analyzed, and some proposals are given for the increase of rural resident revenue.
KeywordsAbsolute error Quasi-stepwise regression Variable selection Rural resident revenue
- 2.Freedman DA (1983) A notes on screening regression equations. Am Stat 3(7):152–155Google Scholar
- 3.Perkins DH (2005) Economic of development, vol 8, issue 4, 5th edn. Renmin University of China Press, Beijing, pp 12–19Google Scholar
- 4.Oh JW (2009) Development of pollen concentration prediction models. J Korean Med Assoc 5(2):579–591Google Scholar
- 5.DN Gujarati (2005) Basic econometrics, vol 9, issue 4, 3rd edn. Renmin University of China Press, Beijing, pp 23–28Google Scholar