昆明市区域商品住宅价格及影响因素研究

发布时间：2018-04-18 07:40

本文选题：商品住宅价格 + 影响因素　；参考：《昆明理工大学》2014年硕士论文

【摘要】：房地产市场最直观的体现就是市场价格,不同因素、不同地区和不同市场对房价影响会有很大的不同,尤其住宅价格与人民生活息息相关。近些年来,房价的不断上涨更是成为人们关注的焦点,学术界也较多选取商品住宅销售均价作为代表商品住房市场发展程度的风向标。研究结合昆明市的房地产市场现状,从众多影响房价的因素中选取了13个指标作为该市房价的代表因素,通过建立灰色关联度模型对尽可能多的因素进行筛选和分析,并辅助MATLAB编程计算出各个因素与房价的绝对和相对关联度值,最终综合考虑确定出每个因素的综合关联度,进而判断其影响大小并排序；依据综合关联度排序的结果,从中选出综合关联系数较大的前七个因素与昆明市房价建立GM(1,8)模型,利用模型对2005-2011年昆明市房价做出预测,并与实际值进行残差检验和后验差检验,证实预测误差较小。在后期预测应用时,虽然各个指标的预测值是由经改进的GM(1,1,x(0))预测获得,但其累积误差不容忽视,为更大程度的提高预测的准确性,随后对构建的GM模型进行改进：调用MATLAB的多项式曲线拟合对各指标进行预测,借助BP神经网络训练和修正其指标预测值的结果,将修正之后的预测值重新代入构建的GM模型,这在很大程度上控制了偏差,得出了更具参考价值和说服力的结论,也弥补了影响房价变化的多个经济变量之间的定量关系无法用精确的数学表达式来描述的不足。最后,对昆明市商品住宅价格进行了短期预测,并依据未来市场波动的不同情形分别予以阐述,结论部分客观的对昆明市区域房价及影响因素做了分析,从宏观调控层面给出了政策建议,此研究为昆明市的投资决策和房地产市场发展提供了一定的指导和借鉴意义。
[Abstract]:The most intuitive embodiment of the real estate market is the market price, different factors, different regions and different markets will have a very different impact on house prices, especially housing prices and people's lives are closely linked.In recent years, the rising of house prices has become the focus of attention, and the academic circles also choose the average price of commercial housing as the vane to represent the development of commodity housing market.According to the present situation of real estate market in Kunming, 13 indexes are selected as the representative factors of housing price from many factors, and the grey correlation model is established to screen and analyze as many factors as possible.The absolute and relative correlation degree between each factor and house price is calculated by assistant MATLAB programming. Finally, the comprehensive correlation degree of each factor is determined synthetically, and then the influence is judged and sorted, and according to the result of comprehensive correlation degree ranking,The first seven factors with large comprehensive correlation coefficient and the housing price in Kunming are selected to establish GM1 / 8) model. The model is used to predict the housing price in Kunming from 2005 to 2011, and the residual error test and posterior error test are carried out with the actual value, which proves that the prediction error is small.In the application of later prediction, although the prediction value of each index is obtained by the improved GM1 / 1X / 0) prediction, the cumulative error can not be ignored, so as to improve the accuracy of the prediction to a greater extent.Then the GM model is improved: the polynomial curve fitting of MATLAB is used to predict each index, and the modified GM model is replaced by BP neural network to train and revise the predicted value of the index.To a great extent, the deviation is controlled, and the conclusion of more reference value and persuasion is drawn, which also makes up for the deficiency that the quantitative relationship between several economic variables that affect the change of house price cannot be described by precise mathematical expression.Finally, the short-term forecast of commodity housing price in Kunming is given, and the different situations of future market fluctuation are expounded respectively. The conclusion is that the regional housing price and its influencing factors in Kunming are analyzed objectively.The policy suggestions are given from the aspect of macro-control, which provides some guidance and reference for the investment decision and the development of real estate market in Kunming.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F299.23

【参考文献】