Hypothesis test of statistical model adequacy in the rotatable experiment design

The elements of multiple regression analysis which are the basis to calculation parameter estimates necessary for constructing the process model are considered. Special plans used in experimental data processing are presented and the least square method applied to the tasks of mathematical models construction is described. The questions of optimal experiment design to construct mathematical models as a linear combination of linear and quadratic functions of input factors with unknown parameters are discussed. Complete and fractional factorial designs, as well as composite orthogonal and rotatable experimental designs for quadratic models are presented. The situations in which a regression model form is unknown exactly to a researcher and is postulated by him are considered. The parameter estimates bias of a postulated model caused by its noncoincidence with the true is studied. The connection between these issues and the ones of general linear hypothesis testing in the model parameters analysis is examined. The methods of important factors to be considered in mathematical models construction are described.

References

[1] Sidnyaev N.I., Mel’nikova Yu.S. Otsenki statisticheskikh parametrov raspredeleniy [Estimates of the Statistical Distribution Parameters]. Moscow, MGTU im. N.E. Baumana, 2012, no. 0321201235. Available at: http://wwwcdl.bmstu.ru/fn1/OcenkiSPR.html (accessed 27.04.2015).

[2] Eliseeva I.I., Yuzbashev M.M. Obshchaya teoriya statistiki [General Theory of Statistics]. Moscow, Finansy i statistika Publ., 2002. 480 p.

[3] Sidnyaev N.I., Vilisova N.T. Vvedenie v teoriyu planirovaniya eksperimenta [Introduction to the theory of experiment planning]. Moscow, MGTU im. N.E. Baumana Publ., 2011. 463 p.

[4] Goryainov V.B. Locally Most Powerful Rank Criteria of Independence of Observations in Model of Spatial Autoregression. Vestn. Mosk. Gos. Tekh. Univ. im. N.E. Baumana, Estestv. Nauki [Herald of the Bauman Moscow State Tech. Univ., Nat. Sci.], 2010, no. 4, pp. 16-28 (in Russ.).

[5] Sidnyaev N.I. Teoriya planirovaniya eksperimenta i analiz statisticheskikh dannykh [The Experimental Design Theory and Analysis of Statistical Data]. Moscow, Yurayt Publ., 2014. 495 p.

[6] Sidnyaev N.I., Sadykhov G.S., Savchenko V.P. Modeli i metody otsenki ostatochnogo resursa izdeliy radioelektroniki [Models and Methods for Assessing Residual Life of Electronics Products]. Moscow, MGTU im. N.E. Baumana Publ., 2015. 382 p.

[7] Gusarov V.M. Teoriya statistiki [Theory of Statistics]. Moscow, YuNITI Publ., 2001. 247 p.

[8] Sidnyaev N.I., Levin V.A., Afonina N.E., Kats A.M. Mathematical Modeling of the Heat-Transfer Intensity by the Methods of the Theory of Experiment Design. Journal of Engineering Physics and Thermophysics, 2002, vol. 75, no. 2, pp. 432-440.

[9] Strizhov V.V. Metody induktivnogo porozhdeniya regressionnykh modeley [Methods of Inductive Generation of Regression Models]. Moscow, Vychislitel’nyy tsentr im. A.A. Dorodnitsyna RAN Publ. [Dorodnitsyn Computing Centre of RAS], 2008. 54 p.

[10] Berezhnaya E.V., Berezhnoy V.I. Matematicheskie metody modelirovaniya ekonomicheskikh sistem [Mathematical Methods of Modeling Economic Systems]. Moscow, Finansy i statistika Publ., 2006. 432 p.

[11] Hastie T., Taylor J., Tibshirani R., Walther G. Forward stagewise regression and the monotone lasso. Electronic Journal of Statistics, 2007, vol. 1, no. 1, pp. 1-29.

[12] Pavlov I.V. Calculation of Some Quality and Reliability Indices for a System with Parallel Strength Members. Jelektr. nauchno-tekh. izd. "Inzhenernyy zhurnal: nauka i innovacii" [El. Sc.-Tech. Publ. "Eng. J.: Science and Innovation"], 2012, iss. 7. Available at: http://engjournal.ru/catalog/mathmodel/technic/296.html DOI: 10.18698/2308-6033-2012-7-296

[13] Bishop C.M., Lasserre J. Generative or discriminative? Getting the best of both worlds. In Bayesian Statistics 8; cd. by J. M. e.a. Bernardo. Oxford University Press, 2007, pp. 3-23.

[14] Strizhov V.V. Search for a parametric regression model in an inductive-generated set. Computational Technologies, 2007, vol. 12, no. 1, pp. 93-102 (in Russ.).

[15] Sidnyaev N.I., Andreytseva K.S. Independence of the Residual Quadratic Sums in the Dispersion Equation with Noncentral x2-Distribution. Applied Mathematics, 2011, vol. 2, по. 10, pp. 1303-1308. Available at: http://file.scirp.org/Html/7855.html DOI: 10.4236/am.2011.210181

[16] Efron B., Hastie T., Johnstone I., Tibshirani R. Least angle regression. The Annals ofStatistics, 2004, vol. 32, no. 3, pp. 407-499.

[17] Gaydyshev I.P. Analiz i obrabotka dannykh: spetsial’nyy spravochnik [Data Analysis and Processing: Special Handbook]. St. Petersburg, Piter Publ., 2001. 752 p.