| Summary: | 碩士 === 國立臺灣科技大學 === 營建工程系 === 96 === This study aimed to apply the hedonic price method to estimate housing prices in Da-an district and Hsinyi district. Used-residential houses at the intermediate-high level were viewed as a segment in housing market segmentation. Through expert interviews, 22 attributes that could reflect the price of used-residential houses at the intermediate-high level were selected. Considering the multicollinearity between attributes of used-residential houses at the intermediate-high level in Da-an and Hsinyi districts, principal component analysis was applied to reduce variables and multicollinearity between attributes, and multiple regression analysis was further conducted to construct a hedonic price model. The advantage of this model was that the attribute variables used were reduced through principal component analysis, so the model would be simpler and less complicated. The drawback was that the reduced attribute variables were derived from the initial attribute variables, so in practical applications, the standardized value of each reduced principal component should be derived according to the standardized value of each initial attribute variable before it could be used in the hedonic price model to obtain the hedonic price. This drawback would increase the difficulty of using the constructed hedonic price model.
The estimation accuracy of the constructed hedonic price model could be further enhanced through improvement of attribute selection, clarification of actual trading price and area. Besides, samples with special conditions could also be excluded to increase the price estimation accuracy of the constructed model.
|
|---|
