On a Generalization of the Modified Gravity Model
The modified gravity model is a nonlinear regression model for estimating passenger correspondences between pairs of spatial points [Andronov and Santalova, Simul. Comput. 41(2), 730–745 (2012)]. Unknown parameters of the model and correspondences are estimated using aggregated data, i.e. total number of passenger departures at each point in a considered time interval. The model estimation under the assumption that the error term is distributed normally is carried out in [Andronov and Santalova, Simul. Comput. 41(2), 730–745 (2012); Santalova, Autom. Control Comput. Sci. 47(2), 99–106 (2013)]. However non-symmetry of the passenger flows may violate the assumption about normality of error distribution. In the present research the model with normally distributed errors is generalized to a model with skew-normal error distribution, and a method of its estimation is proposed.
KeywordsNonlinear Regression Model Passenger Flow Considered Time Interval Standard Normal Density Correspondence Matrix
The research was supported by the European Union through the European Social Fund (Mobilitas grant No GMTMS280MJ) and Estonian Science Foundation (grant No ETF9127).
- 7.Otruzar, J., Willumsen, L.G.: Modeling Transport, 4th edn. Wiley, New York (2011)Google Scholar
- 9.Santalova, D.: Non-symmetrical transport flows estimation using the modified gravity model. In: Mariagiulia M. (eds.) 7th International Workshop on Simulation, pp. 312–313. Department of Statistical Sciences, Unit of Rimini, University of Bologna, Bologna, May 21–25 (2013). (Dipartimento di Scienze Statistiche “Paolo Fortunati”, Alma Mater Studiorum Universita di Bologna, Serie Ricerche)Google Scholar
- 10.Santalova, D.: Migration flows estimation using the modified gravity model. In: Christos H.S. (eds.) 15th Applied Stochastic Models and Data Analysis International Conference, 25–28 June 2013, Mataró (Barcelona), International Society for the Advancement of Science and Technology, pp. 186–186 (2013)Google Scholar