Research on the Application of Integrated RG-LSTM Model in House Price Prediction

Abstract

Aiming at the problem that the traditional time series analysis method cannot fit the nonlinear components in the time series, and the conventional regression analysis cannot reflect the relationship between the time series, it is proposed to use the lag observation method to reconstruct the phase space of the multi-dimensional data, transform the time series problem into a supervised regression problem, and then use the stacking integration idea to establish a regression analysis and long-term and short-term memory network combination model (RG-LSTM), and use it for house price prediction. Firstly, based on the literature review method, the housing price index system is established; secondly, on this basis, the RG-LSTM model is applied to the prediction of housing prices in Nanjing, and comparative experiments are carried out. Finally, the experimental results show that the integrated RG-LSTM model has the advantages of high prediction accuracy and high reliability compared with the traditional time series prediction model, which indicates that the model has great advantages in predicting housing prices.