Composite Quantile Regression And Its Application In Time Sequence

This dissertation studies the estate investment problem by the method ofcomposite quantile regression model and the time series analysis based on the actualproblems about the Chinese real estate investment. The model is set up and forecastresults are arrived based on real estate’s data. First of all, through the simulationcalculation, the parameter estimation, which coming from the method of compositequantile regression, is consistency, and we use the actual data to build model. The mainachievements can be summarized as follows:1. We use the method of composite quantile regression to estimate stationary timeseries model, to get the parameter estimation value. Through this, we can prove theparameter estimation, which coming from the method of composite quantile regressionis consistency, and we use the actual data to build model. We use the2002-2014Chinareal estate investment accumulated growth data，with the method of compositequantile regression and the time series, to set up model and forecast. The regressionmodel is established by the cumulative growth of real estate investment, thecumulative growth of residential real estate investment, the cumulative growth of realestate investment and the office building growth of real estate commercial businessspace investment, and at the last we forecast the result. The results of the study showthe model can objectively describe the relationship of each factor in Chinese real estateinvestment, and the result of forecast is better.2. We use the method of local composite quantile regression to estimate theregression function of the nonparametric model, showing the obtained parameterestimators is consistency. Then, we use the actual data to build model. We use the2001-2014Chinese real estate investment accumulated growth and money supply data,taking use of the method of local composite quantile regression and the time series, tobuild model and forecast the result. The results of the study show the model canobjectively describe the relationship between China real estate investment and moneysupply, and the result of forecasts is better.