Applying Fuzzy Adaptive Networks to Forecasting Real Estate Prices–A Case Study of Nei-Hu and Da-An District Real Estate Market

• Main Authors: Jian-Jiun Chen, 陳建鈞
• Other Authors: Kuen-Tai Chen
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/82881010156954313584

碩士 === 明志科技大學 === 工業工程與管理研究所 === 99 === The main purpose of this paper is to forecast the price of pre-owned house of edifice in Taipei City using Fuzzy Adaptive Networks (FAN), Back Propagation Neural Networks (BPNN), and Adaptive Neuro-Fuzzy Inference System (ANFIS). Except for crisp variables, fuzzy variables are also included in this study namely Vital function, Peripheral environmental condition, and Anticipated development potential. Main steps in this study are: (1) two-step cluster to cluster 12 administrative districts in the Taipei City. (2) Using Da-an District and Nei-hu District as exsamples to conduct a questionnaire survey for obtaining fuzzy variables subjective ratings. (3) Apply FAN, BPNN, and ANFIS to forecast the price of houses in the Taipei City. It was concluded that: (1) The 12 districts in Taipei City could be divided to cluster 1 and 2, according to the proposed two-step clustering. (2) The prediction of Nei-hu District is better than the one of Da-an District. (3) The prediction of FAN is better than BPNN and ANFIS.