Loading...
Please wait, while we are loading the content...
Similar Documents
Avaliaçao de imóveis urbanos baseada em métodos estatísticos multivariados
| Content Provider | Semantic Scholar |
|---|---|
| Author | Alves, Valdir |
| Copyright Year | 2005 |
| Abstract | In this work concepts of Multivariate Statistics are shown and it also presents a computational program denominated MAI (Multivariate Analyses of Immobiles), in MATLAB ®.That program has the purpose to evaluate immobile urban considering each immobile one as a vector of aleatory characteristics and construction of a statistical model that esteems the value of a new property in an automatic way, avoiding appraisers' subjectivities. The program offers to the user three regression options: 1) Clusters Analysis and Main Components applied to the characteristics of the properties for obtaining of homogeneous classes, it also offers 5 (five) types of distances and 4 (four) types of connections (offering the best connection option according to the cophenet index) and those variables were reduced and transformed in way to avoid multicolinearity problems, common to that type of data. Like this, the new variables were used for the determination of the models of Multiple Linear Regression of each class or homogeneous group of items and for each type of properties (apartment marries and land). 2) Factorial Analysis that explains the correlations among the original variables in terms of a group of little you varied you didn't observe, it makes the rotation of the factors (Varimax) to avoid the multicolinearity. 3) Simple Multiple Regression uses all the variables, without any treatment. That option is not reliable, because the data didn't suffer any treatment to avoid multicolinearity problems. The three options present the regression function, the statistical parameters (R 2, F and p) and the esteem value for a new one immobile. The program developed for computer, as well as your evaluation proposal, the was applied a group of data originating from of the research accomplished by Silvia Neide Braulio (2004), referring to 44 apartments, 51 houses and 24 lands of Campo Mourao's city PR, it could be applied at other places and also the other products that need evaluation. The model of each class of the three types of appraised properties presented a good adjustment to the data and a good capacity predictable, assisting all the theoretical suppositions of multiple linear regression. Word-key: evaluation of properties, multiple linear regression, clusters analysis, factorial analysis. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://acervodigital.ufpr.br/bitstream/handle/1884/3234/valdir%20avaliacao.pdf;jsessionid=1B61EE50A655FC29B1505EF6651F8762?sequence=1 |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
