|
Chinese people are most concerned about the current prices , as the object of research topics , with a wide range of topics of significance prices not only people concerned government departments as well as real estate developers are also very concerned about . Currently , the real estate market price forecast is gradually becoming an academic research focus , different theories and methods used , and are trying to explore the internal laws of the real estate market price movements by modeling the price trend analysis and price forecast [ 1 ] . Ningbo is a sub-provincial coastal cities , cities specifically designated in the state plan , the residential real estate market trends in Ningbo typical paper, time series analysis method based on artificial neural network , according to the Real Estate Trading Center in Ningbo City and Ningbo City Real Estate Registration Management Supervision data released at the official website of each month by month , six districts of Ningbo City residential real estate price data , data preprocessing , modeling , learning fitting , to achieve of the Ningbo six residential real estate prices in the short-term forecast , to a certain degree of accuracy . The main work and results are as follows : constructed multivariate time series BP network topology ; 2 . Forecast the dynamic neural network theory is applied to the real estate prices were constructed not long synchronization based on BP neural network , RBF neural network based on time series toolbox of Nnstart GUI residential real estate prices time series forecasting model , the actual data for Ningbo residential real estate simulation prediction experiments showed that the effectiveness of the proposed method ; 3 . experimental results of the three models for horizontal , vertical than analysis , not synchronized long dynamic neural network research ; 4. Summary draw a \
|