Two-stage modelling of random phenomena
cytuj
pobierz pliki
RIS BIB ENDNOTEChoose format
RIS BIB ENDNOTETwo-stage modelling of random phenomena
Publication date: 12.2015
Geoinformatica Polonica, 2015, Vol. 14 (2015), pp. 41 - 49
Authors
Two-stage modelling of random phenomena
The main objective of this publication was to present a two-stage algorithm of modelling random phenomena, based on multidimensional function modelling, on the example of modelling the real estate market for the purpose of real estate valuation and estimation of model parameters of foundations vertical displacements. The fi rst stage of the presented algorithm includes a selection of a suitable form of the function model. In the classical algorithms, based on function modelling, prediction of the dependent variable is its value obtained directly from the model. The better the model refl ects a relationship between the independent variables and their effect on the dependent variable, the more reliable is the model value. In this paper, an algorithm has been proposed which comprises adjustment of the value obtained from the model with a random correction determined from the residuals of the model for these cases which, in a separate analysis, were considered to be the most similar to the object for which we want to model the dependent variable. The effect of applying the developed quantitative procedures for calculating the corrections and qualitative methods to assess the similarity on the fi nal outcome of the prediction and its accuracy, was examined by statistical methods, mainly using appropriate parametric tests of signifi cance. The idea of the presented algorithm has been designed so as to approximate the value of the dependent variable of the studied phenomenon to its value in reality and, at the same time, to have it “smoothed out” by a well fi tted modelling function.
Adamczewski Z. 2002. Nieliniowe i nieklasyczne algorytmy w geodezji. Warszawa: Oficyna Wydawnicza Politechniki Warszawskiej.
Barańska A. 2004. Kryteria stosowania modeli stochastycznych w predykcji rynkowej wartości nieruchomości. Rozprawa doktorska. AGH. Kraków
Barańska A. 2010. Statystyczne metody analizy i weryfikacji proponowanych algorytmów wyceny nieruchomości. Kraków: Wydawnictwa AGH.
Barańska A., Michalik S. 2014. Variants of Modeling Dwelling Market Value. Real Estate Management and Valuation, vol. 22, no. 3, s. 28–35.
Czaja J. 1997. Modele statystyczne w informacji o terenie. Kraków: Wydawnictwa AGH.
Czaja J. 1996. Wybrane zagadnienia z geodezji inżynieryjnej. Kraków: Wydawnictwa AGH.
Czaja J., Parzych P. 2007. Szacowanie rynkowej wartości nieruchomości w aspekcie międzynarodowych standardów wyceny.Kraków: Stowarzyszenie Naukowe im. St. Staszica.
Rao R.C. 1978. Modele liniowe statystyki matematycznej. Warszawa: PWN.
Information: Geoinformatica Polonica, 2015, Vol. 14 (2015), pp. 41 - 49
Article type: Original article
Titles:
Two-stage modelling of random phenomena
Faculty of Mining Surveying and Environmental Engineering,
AGH University of Krakow
Published at: 12.2015
Article status: Open
Licence: None
Percentage share of authors:
Article corrections:
-Publication languages:
EnglishView count: 1535
Number of downloads: 1117