本研究旨在利用機器學習技術中的隨機森林法對臺南市桂田磐古社區的房價進行預測分析。近年來，房地產市場受到多種經濟與政策因素影響，房價波動顯著，傳統的估價方法在處理大量數據與非線性特徵時存在不足。因此，本研究採用隨機森林模型，透過多決策樹的集成學習法，提高預測準確度，並進一步探討影響房價的關鍵因素。
臺灣自 2012 年推行實價登錄制度後，房地產交易資訊的透明度提升，為市場參與者提供了更完整的參考依據，本研究參採內政部實價登錄系統之數據，並結合經濟指標如消費者物價指數 (CPI)、房貸利率與消費者信心指數 (CCI) 進行綜合分析，研究過程首先進行資料清理與前處理，包括刪除異常值、補足缺失數據，並將原始數據轉換為可供模型訓練的結構化格式，再透過變數篩選影響因子，並以歷史的交易資料作為訓練集，建立隨機森林預測模型，最後使用交叉驗證方式分析臺南市桂田磐古社區未來三年的房價走勢。
研究結果顯示，隨機森林模型能有效降低誤差，其中R-Squared值有0.9以上，在預測未來三年房價時能提供更高的準確率與穩定性。本研究不僅驗證了機器學習技術在房價預測的可行性，也為政府政策調整與民眾購房決策提供實證參考。透過更精準的房價預測模型，購房者可更理性評估市場價格，避免因資訊不對稱而產生的交易風險；政策制定者亦可利用此模型評估市場趨勢，調整相關房地產政策，以促進市場穩定發展。

This study employs machine learning techniques, specifically the Random Forest method, to predict housing prices in the Greaten Pangu Community of Tainan City. In recent years, the real estate market has experienced significant price fluctuations due to various economic and policy factors. Traditional valuation methods often struggle to process large datasets and capture nonlinear relationships. To address these limitations, this study utilizes the Random Forest model, which integrates multiple decision trees to enhance predictive accuracy and identify key factors influencing housing prices.
Following the implementation of Taiwan’s Actual Price Registration System in 2012, real estate transaction transparency has improved, providing a more comprehensive data source for market analysis. This study incorporates transaction data from the Ministry of the Interior, along with economic indicators such as the Consumer Price Index (CPI), mortgage interest rates, and the Consumer Confidence Index (CCI). The research process includes data cleaning, preprocessing, variable selection, and model training using historical transaction records. Cross-validation is then applied to forecast housing price trends for the next three years.
Results indicate that the Random Forest model effectively reduces prediction errors, achieving an R-squared value above 0.9, demonstrating high accuracy and stability. This study confirms the feasibility of machine learning in housing price prediction and provides valuable insights for policy adjustments and homebuyer decision-making, contributing to a more transparent and stable real estate market.

1.Andrew T. Court,“Hedonic Price Indexes with Automotive Examples.”, Automobile Manufacturers Association, 1939.
2.Alican Dogan,Derya Birant,“Machine Learning and Data Mining in Manufacturing.”, Expert Systems with Applications, Volume 166, 2021
3.Peter Cabena, Pablo Hadjinian, Rolf Stadler, Jaap Verhees, and Alessandro Zanasi, “Discovering Data Mining: From Concept to Implementation.”, Prentice-Hall, Inc., United States, 1998
4.Felipe D. Calainho1,Alex M. van de Minne, Marc K. Francke1,“A Machine Learning Approach to Price Indices: Applications in Commercial Real Estate.”, The Journal of Real Estate Finance and Economics, Volume 68, 624-653, 2024
5.George C. Haas,“Sale Prices as a Basis for Farmland Appraisal.”, University Farm, 1922a.
6.George C. Haas,“A Statistical Analysis of Farm Sales in Blue Earth County, Minnesota, as a Basis for Farm Land Appraisal.”, Technical report of University of Minnesota, 1922b.
7.G. Donald Jud and Daniel T. Winkler,“The Dynamics of Metropolitan Housing Prices.”, Journal of Real Estate Research, Volume 23, 29-46, 2002
8.Henry A. Wallace,“Comparative Farmland Values in Iowa.”, Journal of Land and Public Utility Economics, 385-392, 1926.
9.James P. LeSage, R. Kelley Pace,“Models for Spatially Dependent Missing Data.”, The Journal of Real Estate Finance and Economics, Volume 29, 233-254, 2004
10.Jengei Hong, Heeyoul Choi, Woo-Sung Kim,“A house price valuation based on the random forest approach: the mass appraisal of residential property in South Korea.”,The International Journal of Strategic Property Management, Volume 24 No.3, 2020
11.Kelvin J. Lancaster,“A new approach to consumer theory.”, Journal of Political Economy, 74(2), 132-157, 1966.
12.Leo Breiman,“Random Forests.”, Machine Learning, Volume 45, 5-32, 2001
13.Roger E. Cannaday, Peter F. Colwell,“The Cost Approach: Appraisal Gone Wrong?”, Illinois Real Estate Letter, 1995
14.Sherwin Rosen,“Hedonic prices and implicit markets: product differentiation in pure competition.”, Journal of Political Economy, 82(1), 34–55, 1974.
15.Sunita M. Dol, Pradip M. Jawandhiya,“Classification Technique and its Combination with Clustering and Association Rule Mining in Educational Data Mining — A survey.”, Engineering Applications of Artificial Intelligence, Volume 122, 2023.
16.Tingting Chen, Shijing Si,“Predicting Rental Price of Lane Houses in Shanghai with Machine Learning Methods and Large Language Models.”, arXiv, 2024.
17.Walter Coleman, Ben Johann, Nicholas Pasternak, Jaya Vellayan, Natasha Foutz, Heman Shakeri, “Using Machine Learning to Evaluate Real Estate Prices Using Location Big Data.”, arXiv, 2022.
18.Zvi Griliches,“Hedonic price indexes for automobiles: An econometric of quality change.”, In The price statistics of the federal goverment., NBER, 173–196, 1961.
19.王洺方。以多變量方法以及機器學習模型預測不動產價格－以台中市實價登錄資料為例。碩士論文。國立中興大學。2024。
20.朱正中。實價登錄制度對不動產經紀業者的影響與挑戰。碩士論文。國立中正大學。2024。
21.邱若綺。商業綜合體對周邊房價之影響--以台中勤美誠品綠園道為例。碩士論文。國立中興大學。2024。
22.洪暉恒。台北市大型商用建築物成本法估價公平性評估-估價比值應用。碩士論文。國立政治大學。2020。
23.許智淵。類神經網路技術與特徵價格法於台北市房價預測結果之比較研究。碩士論文。國立臺北大學。2012。
24.崔淑琄。總體經濟對台灣都會區域房地產市場影響。碩士論文。國立臺北大學。2020。
25.黃允亭。應用實價登錄建立以聚類方法之堆疊泛化房價預測模型-以桃園市區分建物房價資料為例。碩士論文。國立政治大學。2022。
26.楊鈞喬。房市泡沫預測：長時期的總體經濟歷史資料和隨機森林演算法能夠告訴我們甚麼？。碩士論文。國立臺灣大學。2021。
27.劉富容、游璿達、黃孝雲、劉正夫。利用政府開放資料探討影響台北市房價之主要房屋特性及周邊設施影響因子。Journal of Data Analysis，14卷，5期，1-26。2019。
28.內政部不動產交易實價查詢服務網. (n.d.). Available from https://lvr.land.moi.gov.tw/
29.中華民國統計資訊網. (n.d.). Available from https://www.stat.gov.tw/Default.aspx
30.臺灣經濟發展研究中心. (n.d.). Available from http://rcted.ncu.edu.tw/index.asp
31.中華民國中央銀行全球資訊網. (n.d.). Available from https://www.cbc.gov.tw/tw/mp-1.html