APPLICATION OF FUZZY TIME SERIES FOR PREDICTING REAL ESTATE PRICES IN TAIWAN

Real estate has always been the basic demand of settlement and the optimal hedge option of investment and financial management for humanity. It is also the leading industry that drives national economic development while exerting enormous influence on the industrial economic development of the country. Consequently, being able to predict the fluctuation in real estate prices is one of the most important topics for government policy making, industrial investment, and individual commerce. This study established a forecast model of real estate prices by using the analytic method of fuzzy time series. The data source was the time series analytic statistics of the sale contract average unit price registered on the real estate information platform of the Ministry of the Interior. Overall real estate price predictions in Taiwan, especially those for three major cities in northern, central, and southern Taiwan were made. The analysis and prediction results showed that the fuzzy time series forecast model applied in this study had error percentages lower than 5% in the annual and seasonal prediction analytic results. Thus, the forecast model has considerably high prediction accuracy.