The paper discusses a synthesis of multiple linear regression and Deming regression model’s, which gives countless different and previously unknown estimates of unknown parameters in framework of ordinary least squares. Using example of railway freight turnover modeling for synthesized regression, for the first time, dependences of parameter estimates and adequacy criteria on a ratio of error variances of variables are obtained. It is shown that using the considered synthesis it is possible to reduce the approximation qualities of the classical multiple linear regression model for the sake of improving some of its other important characteristics. Also, the synthesized model can act as a tool for solving the problem of subset selection in regression.
1. Ajvazyan S.A., Enyukov I.S., Meshalkin L.D. Prikladnaya statistika: Issledovanie zavisimostej [Applied Statistics: Addiction Research]. Moscow, Finance and Statistics, 1985, 487 p.
2. Draper N.R., Smith H. Applied regression Analysis, 3rd edition. John Wiley & Sons, 1998, 736 p.
3. Dougerti K. Vvedenie v ehkonometriku [Introduction in Econometrics]. Moscow, Infra-M, 2009, 465 p.
4. Deming W.E. Statistical adjustment of data. New York: Wiley, 1943, 273 p.
5. Bazilevskij M.P. Analiticheskie zavisimosti mezhdu koehfficientami determinacii i sootnosheniem dispersij oshibok issleduemyh priznakov v modeli regressii Deminga [Analytical dependencies between the coefficients of determination and the ratio of the error dispersions of the studied features in the Deming regression model]. Matematicheskoe modelirovanie i chislennye metody [Mathematical modeling and numerical methods]. 2016, vol.10, no. 2, pp. 104–116.
6. Bazilevskij M.P. Analiticheskie zavisimosti dlya nekotoryh kriteriev adekvatnosti modeli regressii Deminga [Analytical dependencies for some criteria of adequacy of the Deming regression model]. Vestnik IrGTU [Bulletin of ISTU]. Irkutsk, 2016, vol. 20, no. 10, pp. 81–89.
7. Bazilevskij M.P. Sintez modeli mnozhestvennoj linejnoj regressii i regressii Deminga [Synthesis of the Multiple Linear Regression and Deming Regression Model]. Informacionnye tekhnologii v modelirovanii i upravlenii: podhody, metody, resheniya: materialy II Vserossijskoj nauchnoj konferencii s mezhdunarodnym uchastiem [Information technologies in modeling and management: approaches, methods, solutions: materials of the II All-Russian Scientific Conference with international participation]. Tolyatti, 2019, part 1, pp. 64–69.
8. Noskov S.I. Tehnologija modelirovanija ob’ektov s nestabil'nym funkcionirovaniem i neopredelennost'ju v dannyh [Modeling technology for objects with unstable operation and data uncertainty]. Irkutsk, RIC GP «Oblinformpechat'» Publ., 1996, 321 р.
9. Noskov S.I., Bazilevskij M.P. Postroenie regressionnyh modelej s ispol'zovaniem apparat linejno-bulevogo programmirovaniya [Construction of regression models using linear-boolean programming device]. Irkutsk, IrGUPS, 2018, 176 p.