本文應用「遞迴分區迴歸樹模型」（Model-based recursive partitioning、MOB）來分析影響房價的關鍵因素，並建構房價預測模型。MOB係迴歸與決策樹的混合模型，對於已知是線性或非線性的影響變數以線性或非線性迴歸模型處理之；對於影響非常複雜的變數則納入決策樹中，作為區分變數。本文使用的資料庫是不動產交易實價查詢，這是一個登錄正確性極高且交易的細節皆很完整的資料庫。該資料庫除了房價及房屋基本資訊外，額外的加入「與最近捷運站的距離」、「與最近公園的距離」、「公園的面積」，「移轉樓層」，「使用目地」，「經緯度」及「交易年份」等變數。這些變數都對房價有很大的影響且相當完整，其中某些變數適合以母數迴歸模型處理，如房屋的面積等；有些變數如經緯度則宜以納入決策樹中作為分區變數。混合模型能夠兼顧估計效率，具有非常高的彈性，且適宜處理連續與間斷型值性變數，實證結果發現MOB對於房價有很好的預測能力。
除了「遞迴分區迴歸樹模型」之外，為了進行模型效果的比較，本文額外地加入大量的機器學習相關模型作為比較，其中涵蓋監督式模型以及非監督式模型，部分僅能輸出分類之估計，其餘則能夠輸出連續的估計值。因為分類主要使用房價單位面積價格取自然對數作為分類，因此若僅只能輸出分類之估計的模型，整體表現會較能輸出連續的估計值之模型為佳。涵蓋模型有複迴歸分析、迴歸樹分析、遞迴分區迴歸樹模型、adabag、adaBoost、BlackBoostModel、C50Model、EarthModel、FDAModel、GBMModel、GLMBoostModel、GLMModel、GLMNetModel、KNNModel、LARSModel、LDAModel、LMModel、MDAModel、NNetModel、PLSModel、RangerModel、RpartModel。除前三模型之外，後十九個模型皆使用R的套件MachineShop中提供的機器學習模型，而全部二十二個模型皆是使用R之套件。
使用的模型評價指標有RMSE、MAE、MAPE三項，就全部結果而言，表現最高是隨機森林模型，而遞迴分區迴歸樹模型在其中也算是排名較前的模型。總結這些模型在不動產交易實價查詢該資料庫中之表現，正確地選用機器學型模型能夠提供不同變數對不動產價格的估價，這樣的估價模式能夠讓交易者對價格高低有一個評斷的標準，最終祁使能促使市場交易中的效率性得以提升。

This thesis applies the “Model-based recursive partitioning” (MOB) to analyze the key factors affecting housing prices and construct a housing price prediction model. MOB is a mixed model consisting of tree model and regression model. Those factors with known linear or nonlinear effects are included in the regression model as regressors while those factors with extremely complicated nonlinear impacts are treated as tree partitioning variables. We apply the MOB to the “actual selling price of real estate”, which is a database with accurate registration and complete transaction details. In addition to the basic information of house prices, the database adds “distance to the nearest MRT station”, “distance to the nearest park”, “area of the park”, “transaction floor”, “purpose of usage “, and variables such as “latitude and longitude” and “transaction year”. These variables have a great impact on housing prices and are quite complete. Some of these variables are suitable for processing with a parametric regression model, such as the area of houses, etc. Some other variables, such as latitude and longitude, should be included in the decision tree as partition variables. The hybrid model could achieve estimation efficiency, possess a very high flexibility, and is suitable for dealing with both continuous and discrete variables. Empirical analysis shows that MOB has a good predictive ability for housing prices.
In addition to the “recursive partition regression tree model”, we adds a large number of supervised and unsupervised machine learning models for the purpose of comparison. While some of them can only output categorical estimates, some others can produce continuous estimates. We mainly uses the natural logarithm of the price per unit area of the house price as the output variable. Models under investigation include multiple regression analysis, regression tree analysis, recursive partition regression tree model, adabag, adaBoost, BlackBoostModel, C50Model, EarthModel, FDAModel, GBMModel, GLMBoostModel, GLMModel, GLMNetModel, KNNModel, LARSModel, LDAModel, LMModel, MDAModel, NNetModel, PLSModel , RangerModel, and RpartModel. All of the last nineteen models use the machine learning models provided in the R package MachineShop while the first three use different R packages. The model evaluation criterion used are RMSE, MAE, and MAPE. Empirical analysis finds that the random forest model has the best performance, immediately followed by recursive partition regression tree model. Our analysis sheds lights on selecting the right machine learning model and predictors to evaluate the price of real estate. Our analysis could allow traders to have a sensible evaluation model for the real estate price, and could improve the market efficiency in the real estate market.

第壹章 緒論 1
第一節 研究動機與目的 1
第二節 研究資料 1
第貳章 文獻探討 5
第參章 研究方法 7
第肆章 實證分析 19
第伍章 結論與未來發展 31
第一節 結論 31
第二節 未來發展 32
參考文獻 33

沈育生,林秋瑾. (2012). “不同人工神經網路架構在不動產大量估價之應用與比較” 臺灣土地研究, 15(1), 1-29.

黃耀賢, 林惠敏, 陳奕潔,林永鑫. (2018). “複迴歸分析法與不動產實價登錄資料的台灣房價估算模型之研究” 創新與管理, 14(1), 1-25.

鄒克萬, 張秀玲,張曜麟. (2002). “整合空間統計技術之土地大量估價方法之研究” 都市與計劃, 29(3), 395-420.

Agostini, C. A., & Palmucci, G. A. (2008). “The anticipated capitalisation effect of a new metro line on housing prices” Fiscal studies, 29(2), 233-256.

Anderson, L. M., & Cordell, H. K. (1988). “Influence of trees on residential property values in Athens, Georgia (USA): A survey based on actual sales prices” Landscape and urban planning, 15(1-2), 153-164.

Antipov, E. A., & Pokryshevskaya, E. B. (2012). “Mass appraisal of residential apartments: An application of Random forest for valuation and a CART-based approach for model diagnostics” Expert Systems with Applications, 39(2), 1772-1778.

Din, A., Hoesli, M., & Bender, A. (2001). “Environmental variables and real estate prices” Urban studies, 38(11), 1989-2000.

Dombrow, J., Rodriguez, M., & Sirmans, C. F. (2000). “The market value of mature trees in single-family housing markets” Appraisal Journal, 68(1), 39-43.

Fotheringham, A. S., Charlton, M. E., & Brunsdon, C. (1998). “Geographically weighted regression: a natural evolution of the expansion method for spatial data analysis” Environment and planning A, 30(11), 1905-1927.

François, D. R., Marius, T., Yan, K., & Paul, V. (2002). “Landscaping and house values: an empirical investigation” Journal of real estate research, 23(1-2), 139-162.

Goodman, A. C., & Thibodeau, T. G. (2003). “Housing market segmentation and hedonic prediction accuracy” Journal of Housing Economics, 12(3), 181-201.

Gu, J., Zhu, M., & Jiang, L. (2011). “Housing price forecasting based on genetic algorithm and support vector machine” Expert Systems with Applications, 38(4), 3383-3386.

Guan, J., Zurada, J., & Levitan, A. (2008). “An adaptive neuro-fuzzy inference system based approach to real estate property assessment” Journal of Real Estate Research, 30(4), 395-422.

Han, J., Pei, J., Kamber, M. (2011). “Data Mining: concepts and techniques” Elsevier,Waltham, MA 02451, USA.

Huang, B., Wu, B., & Barry, M. (2010). “Geographically and temporally weighted regression for modeling spatio-temporal variation in house prices” International Journal of Geographical Information Science, 24(3), 383-401.

Lasota, T., Telec, Z., Trawiński, B., & Trawiński, K. (2009, October). “A multi-agent system to assist with real estate appraisals using bagging ensembles” International Conference on Computational Collective Intelligence (pp. 813-824). Springer, Berlin, Heidelberg.

Laverne, R. J., & Winson-Geideman, K. (2003). “The influence of trees and landscaping on rental rates at office buildings” Journal of Arboriculture, 29(5), 281-290.

Lockwood, L. J., & Rutherford, R. C. (1996). “Determinants of industrial property value” Real Estate Economics, 24(2), 257-272.

Lu, B., Charlton, M., Harris, P., & Fotheringham, A. S. (2014). “Geographically weighted regression with a non-Euclidean distance metric: a case study using hedonic house price data” International Journal of Geographical Information Science, 28(4), 660-681.

More, T. A., Stevens, T., & Allen, P. G. (1988). “Valuation of urban parks” Landscape and urban planning, 15(1-2), 139-152.

Nesticò, A., & Bencardino, M. (2016). “Urban real estate values on vast area and macroeconomic parameters” Procedia-Social and Behavioral Sciences, 223, 410-415.

Nghiep, N., & Al, C. (2001). “Predicting housing value: A comparison of multiple regression analysis and artificial neural networks” Journal of real estate research, 22(3), 313-336.

Park, B., & Bae, J. K. (2015). “Using machine learning algorithms for housing price prediction: The case of Fairfax County, Virginia housing data” Expert Systems with Applications, 42(6), 2928-2934.

Schulz, R., & Werwatz, A. (2004). “A state space model for Berlin house prices: Estimation and economic interpretation” The Journal of Real Estate Finance and Economics, 28(1), 37-57.

Selim, H. (2009). “Determinants of house prices in Turkey: Hedonic regression versus artificial neural network” Expert systems with Applications, 36(2), 2843-2852.

Sirmans, S., Macpherson, D., & Zietz, E. (2005). “The composition of hedonic pricing models” Journal of real estate literature, 13(1), 1-44.

Smith, T. E., & Wu, P. (2009). “A spatio-temporal model of housing prices based on individual sales transactions over time” Journal of Geographical Systems, 11(4), 333.

Sun, H., Tu, Y., & Yu, S. M. (2005). “A spatio-temporal autoregressive model for multi-unit residential market analysis” The Journal of Real Estate Finance and Economics, 31(2), 155-187.

Tajani, F., Morano, P., Locurcio, M., & Torre, C. M. (2016). “Data-driven techniques for mass appraisals. Applications to the residential market of the city of Bari (Italy) “ International Journal of Business Intelligence and Data Mining, 11(2), 109-129.

Tyrväinen, L., & Miettinen, A. (2000). “Property prices and urban forest amenities” Journal of environmental economics and management, 39(2), 205-223.
Wu, B., Li, R., & Huang, B. (2014). “A geographically and temporally weighted autoregressive model with application to housing prices” International Journal of Geographical Information Science, 28(5), 1186-1204.

Zeileis, A., Hothorn, T., & Hornik, K. (2008). “Model-based recursive partitioning” Journal of Computational and Graphical Statistics, 17(2), 492-514.

Zhang, R., Du, Q., Geng, J., Liu, B., & Huang, Y. (2015). “An improved spatial error model for the mass appraisal of commercial real estate based on spatial analysis: Shenzhen as a case study” Habitat International, 46, 196-205.