Application of Huber and Hampel M-estimation in Analysing of Real Estate Price Volatility Over Time
Streszczenie
Determining market value of the property requires a thorough and complete analysis of the real estate market. Analysis of the market comes down, inter alia, to determine the state of the real estate market at the time. The law doesn't limit the period of market analysis, provided that property prices will be adjusted on the valuation date or the date of analysis. Obligation to update prices follows from the Act on real estate management. Different methods of determining of the influence of time on real estate prices can be found in the literature. For small data sets can be used interval method, while for large data sets statistical ethods are applicable. Among the statistical methods most frequently cited are regression models. Estimation of linear regression model parameters is performed frequently by using least square method, which is not resistant to outlier cases. Even a single outlier can have a negative effect on the results of the estimation. Alternatively, the model parameters estimation can be made of robust estimation methods. This work presents some robust estimation methods in relation to determining of the influence of time on real estate prices. Parameters estimation results for linear models were compared. Summarizes the results of the estimation of the least squares method and some robust estimation methods. variance analysis being also taken as a basis for conclusion. Analysis and calculations have been carried out on the sample database of properties.
Kolekcje
Korzystanie z tego materiału jest możliwe zgodnie z właściwymi przepisami o dozwolonym użytku lub o innych wyjątkach przewidzianych w przepisach prawa, a korzystanie w szerszym zakresie wymaga uzyskania zgody uprawnionego.