透過您的圖書館登入
IP:3.142.124.252
  • 期刊

房屋價格預測模型分析-以高雄市區爲例

Some Prediction Model for Housing Price in Kaohsiung Area

摘要


民眾在買屋時、賣屋時最想獲得的資訊爲成交行情。房地產投資是一般民眾非常關心的課題,如何以適當的價格找到適合的房屋,是一件不容易的事情,僅憑經驗或一般報導,無法滲入了解狀況,深入分析是有必要的。 本文主要以高雄市區九個行政區域由房仲業者提供的房屋成交價格資料爲基,結合決策樹(Decision Tree)建構模型之分析,研究模型與變數的選擇問題。將成對變數關係利用CART(Classification and Regression Trees)的分析工具,逐次搜索各種屬性組合,及變數水準的合併。如此可降低模型中因子之水準數目,適切掌握變數間的相關與互動,更進一步作資料分析並建構迴歸模型後再精簡變數,使迴歸模型能更有效反應母體的結構。 以往在研究房屋價格因限於成交價格不易取得,通常只討論預售屋、新成屋,而忽略中古屋行情分析;或者選用房價指數,以致考慮的樣本資料過少、時間太長,對房價短期的調整無法掌握,容易造成偏差的情形。本研究乃以高雄都會區房屋之實際交易價格爲基礎,以決策樹呈現房屋價格之結構,並進一步找到預測模型,以瞭解未來房價之變化及建構迴歸模型。

關鍵字

決策樹 預測模型 房價

並列摘要


The investment on real estate is the subject that people concern about very much. The types of housing and the locations are the variables that housing reflects. On the basis the traded prices, this article reveals the structure of housing prices as decision tree show. Moreover, the prediction model tells how the housing prices will change. This paper makes a deal of the price data as the base with Kaohsjung urban area. We combine the Decision tree to construct the prediction model. This research is as the foundation with the real trade price of the house in Kaohsjung area. We presented the structure of the housing price with the decision tree. In order to understand the change of room rate in the future and construct the regression model.

並列關鍵字

Decision tree Prediction model Housing price

延伸閱讀