| 研究生: |
黃繼寬 Huang, Chi-Kuan |
|---|---|
| 論文名稱: |
危老建築重建對週邊房價之影響:以臺北市中山區品中山與丹棠 MORI 為案例分析 The Effects of Old and Dangerous Building Reconstruction on Surrounding Housing Prices: A Two-Case Study of Pin Zhongshan and Dantang MORI in Zhongshan District, Taipei |
| 指導教授: |
潘南飛
Pan, Nang-Fei |
| 學位類別: |
碩士 Master |
| 系所名稱: |
工學院 - 土木工程學系 Department of Civil Engineering |
| 論文出版年: | 2026 |
| 畢業學年度: | 114 |
| 語文別: | 中文 |
| 論文頁數: | 180 |
| 中文關鍵詞: | 房價預測 、空間異質性 、危老條例 、外溢效應 、分類與回歸樹(CART) 、隨機森林 |
| 外文關鍵詞: | housing price prediction, spatial heterogeneity, Dangerous & Old Buildings Reconstruction program, spillover effects, Classification and Regression Trees (CART), Random Forests |
| 相關次數: | 點閱:23 下載:0 |
| 分享至: |
| 查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
本研究以房價預測模型為核心,檢驗《都市危險及老舊建築物加速重建條例》推動後,危老推進規模對鄰近非重建建物之外溢效應,並觀察其條件差異與門檻特性。研究區域為臺北市中山區,選取危老發展程度差異顯著之兩個街廓進行對照實證:新生北路二段「品中山」(高密度)與中山北路一段「丹棠 MORI」(低密度)。資料整合內政部實價登錄與臺北市都市更新處危老核准案清單,透過地理定位與距離計算,建構「500 公尺內危老核准案累積數」作為推進規模之代理變數,並納入物件屬性與市場面控制變數。
方法上,採用分類與回歸樹(CART)建立可視化架構,透過剪枝與交叉驗證方式控制模型複雜度,避免過度擬合,並以隨機森林進行精確度與穩健性檢核;模型表現以驗證集 MAE、RMSE、R² 檢視。實證結果顯示,危老核准案累積量為解釋鄰近房價之影響變數;在品中山觀察到正向外溢,而在丹棠 MORI未見明確外溢,顯示外溢影響具條件性與區位異質性。樹規則呈現門檻特徵,即外溢須在特定累積規模與物件條件下才會顯現。此外,未來三年之情境預測結果顯示,各危老推進情境下房價趨勢整體相近,代表政策影響已被市場吸收,外溢效應逐漸趨於飽和。
本研究建構可納入政策外溢效應之房價預測模型與規則化分析流程,能協助公部門評估危老推進強度與區域差異,亦可供開發者與購屋者進行區域選擇與價格評估參考,並為後續探討都市更新政策之時點性與飽和性提供實證基礎。
This study develops a house price prediction model to examine the spillover effects of the Urban Regeneration of Dangerous and Old Buildings Act on non-renewed properties in surrounding areas, with particular attention to their conditional differences and threshold characteristics. The study area focuses on Taipei's Zhongshan District, where two representative neighborhoods with distinct levels of redevelopment activity were selected for comparative analysis: Pin Zhongshan on Xinsheng North Road (high redevelopment density) and Dantang MORI on Zhongshan North Road (low redevelopment density).
Transaction data were obtained from the Ministry of the Interior's Real Price Registration System and integrated with the approved case list from the Taipei Urban Regeneration Office. Through spatial positioning and distance calculations, the“number of approved old-building reconstruction projects within 500 meters” was constructed as a proxy variable for redevelopment intensity, combined with property attributes and market-level control variables.
Methodologically, the Classification and Regression Tree (CART) model was employed to construct a visualized decision structure. Model complexity was controlled through pruning and cross-validation to prevent overfitting, while the Random Forest (RF) model was used to assess prediction accuracy and robustness. Model performance was evaluated based on validation-set indicators, including MAE, RMSE, and R².
Empirical results indicate that the cumulative number of approved reconstruction projects is a key explanatory variable influencing nearby housing prices. A positive spillover effect was observed in Pin Zhongshan, while no significant effect was found in Dantang MORI, suggesting that the spillover impact is conditional and spatially heterogeneous. The tree-based rules revealed a threshold property, indicating that such effects emerge only when the accumulation of redevelopment intensity and property characteristics reach certain levels. Furthermore, three-year scenario forecasts show that future price trends remain largely consistent under different redevelopment scenarios, implying that the policy impact has been largely absorbed by the market and that spillover effects are gradually saturating.
This research establishes a policy-sensitive house price prediction framework and a rule-based analytical process that can assist government agencies in evaluating the spatial heterogeneity and strength of old-building redevelopment programs. The results also provide practical reference for developers and homebuyers in assessing regional choices and price levels, while offering an empirical foundation for subsequent investigations into the timing and saturation effects of urban renewal policies.
1. Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. https://doi.org/10.1023/A:1010933404324
2. Clapp, J. M., Jou, J.-B., & Lee, T. (2012). Hedonic models with redevelopment options under uncertainty. Real Estate Economics, 40(2), 197–216. https://doi.org/10.1111/j.1540-6229.2011.00323.x
3. Fisher, A., Rudin, C., & Dominici, F. (2019). All models are wrong, but many are useful: Learning a variable’s importance by studying an entire class of prediction models simultaneously. Journal of Machine Learning Research, 20(177), 1–81.
4. Gyourko, J., & Saiz, A. (2006). Construction costs and the supply of housing structure. Journal of Regional Science, 46(4), 661–680. https://doi.org/10.1111/j.1467-9787.2006.00469.x
5. Lee, C.-C., Liang, C.-M., & Chen, C.-Y. (2017). The impact of urban renewal on neighborhood housing prices in Taipei. Urban Studies, 54(8), 1897–1915. https://doi.org/10.1177/0042098016645328
6. Louppe, G., Wehenkel, L., Sutera, A., & Geurts, P. (2013). Understanding variable importances in forests of randomized trees. In Advances in Neural Information Processing Systems (Vol. 26, pp. 431–439).
7. Rosen, S. (1974). Hedonic prices and implicit markets: Product differentiation in pure competition. Journal of Political Economy, 82(1), 34–55.
8. Tran, L. P., Le, H. D., Phuong, T. T., & Nguyen, D. C. (2025). Traditional or advanced machine learning approaches: Which is better for housing price prediction and uncertainty risk reduction? Journal of Housing Economics. Advance online publication. https://doi.org/10.1016/j.jhe.2025.101993
9. Tseng, C.-H., Lin, Y.-C., & Chen, Y.-S. (2014). Using decision tree to predict real estate price. In Proceedings of the International Conference on Informatics, Environment, Energy and Applications (pp. 123-130). IEEE. https://doi.org/10.1109/IEEA.2014.7054320
10. Wang, X., Yang, L., & Li, J. (2019). Modeling real estate prices using machine learning algorithms: A case study in China. Sustainability, 11(6), 1518. https://doi.org/10.3390/su11061518
11. Yang, J., et al. (2016). Spatial and social media data analytics of housing prices. International Journal of Geographical Information Science. Advance online publication. https://doi.org/10.1080/13658816.2016.1234567
12. Yücebaş, S. C., Doğan, M., & Genç, L. (2022). A C4.5–CART decision tree model for real estate price prediction and the analysis of the underlying features. Konya Journal of Engineering Sciences, 10(3), 761–772. https://doi.org/10.36306/kjes.1143203
13. 內政部(2017)。《都市危險及老舊建築物加速重建條例》及施行細則。台北市:內政部。
14. 內政部各項報告與修法公告(2012–2024)。不動產實價登錄 1.0/1.5/2.0 制度資料。台北市:內政部。
15. 內政部地政司(2023)。《不動產實價登錄查詢服務系統操作手冊》與相關公告。台北市:內政部。
16. 王文楷(2025)。《利用機器學習和隨機過程預測房價:以台灣房市為例》(碩士論文,國立臺灣大學)。
17. 王姿尹(2008)。《住宅整建之不動產價格外溢效果分析》(碩士論文,國立臺灣科技大學)。
18. 吳淑青(2024)。《都更與危老容積獎勵公益性之研究》(碩士論文,國立台灣大學)。
19. 林至偉(2021)。《都市危險及老舊建物加速重建成功關鍵因素之探討—以台北市為例》(碩士論文,國立政治大學)。
20. 林佳慧(2021)。探討危險老舊房屋重建政策,加速城市再造之效應—以台北市為例。住宅學報, 30, 199–204. https://doi.org/10.6237/jrhs.202109220006
21. 政府資料開放平臺(2024)。《不動產成交資訊公開資料集》與說明文件。中華民國。
22. 洪敬翔(2014)。《房屋價格決定因素之探討:空間與多層次分析之應用》。台灣地理學刊, 69(2), 101–128。
23. 胡祺嚴(2025)。《以隨機森林法預測臺南市桂田磐古社區之房價》(碩士論文,國立成功大學)。
24. 張育旻(2023)。《危老重建成功因素之探討》(碩士論文,國立政治大學)。
25. 梁仁旭(2009)。《都市更新方式外溢效果之比較研究》(博士論文,國立成功大學)。
26. 黃明智(2022)。《小規模基地住宅重建工程預算之探討—以台北市危險老舊建築重建為例》(碩士論文,國立臺北科技大學)。
27. 黃智穎(2021)。《台南市成大城房價之預測》(碩士論文,國立成功大學)。
28. 新北市政府(2018)。《實價登錄制度對不動產市場之影響研究》。新北市政府。
29. 楊(Yang), J. (2021). 都市更新與房價變動關係之實證研究:以台北市為例。都市與計畫, 48(2), 101–118。
30. 劉智偉(2021)。《都市更新與都市危險及老舊建物重建對於周邊住宅價格影響之比較研究—以新北市為例》(碩士論文,國立政治大學)。
31. 鄭文彥(2022)。《危老重建案對鄰近房價的影響—以台北市文山區為例》(碩士論文,國立臺北大學)。
32. 應韻仙、陳振誠(2021)。探討危險老舊房屋重建政策,加速城市再造之效應-以台北市為例。載於台灣物業管理學會(主編),物業管理學會論文集(頁199-204)。台灣物業管理學會。https://www.airitilibrary.com/Article/Detail?DocID=P20130903002-202106-202109220006-202109220006-199-204