基于BP神经网络的我国房地产市场风险评价研究

发布时间：2018-02-13 09:36

本文关键词： 房地产市场风险评价 BP神经网络　出处：《湘潭大学》2014年硕士论文　论文类型：学位论文

【摘要】：在我国，，房地产业在国民经济中占有举足轻重的地位，不仅关系着人民的生活水平和住房保障，同时也影响着每个地区乃至整个国家的经济体系稳定程度。近些年来，由我国房地产市场出现的种种现象反应，我国房地产市场存在着不容忽视的风险。市场上也频繁出现一些关于房地产泡沫过高，房地产风险过大的言论。然而房地产市场自身所具有的非线性等特征，使得对其风险程度的评价变得相对困难。本文正是在此背景之下，针对房地产的非线性特征，构建了完整、科学的评价指标体系，通过神经网络模型来准确的评估我国房地产市场存在的风险程度。本文根据房地产市场的发展度以及和谐度一共12个指标构建了我国房地产市场的评价指标体系，并通过BP神经网络模型针对我国31个省及直辖市的2001年至2012年的房地产样本数据进行建模并预测分析。借助了Matlab7.0中人工神经网络模块，实现BP神经网络模型的建立，并经过检测，发现模型具有良好的泛化能力。最后利用建立的模型对我国31个省及直辖市2013年的房地产市场风险情况进行了预测分析，得出我国房地产风险程度及分布情况：（1）我国房地产市场整体风险较大，31个省及直辖市中只有5个处于风险较小的情况；（2）我国房地产市场风险从东部沿海发达地区由大到小向西部欠发达地区分布。最后，针对我国房地产市场风险状况，从经济手段以及法律手段两个方面给出了对策建议。本文的主要创新在于：根据房地产市场的发展度和和谐度构建了全面、合理的房地产市场风险评价指标体系；同时利用BP神经网络模型从整个国家层面分省市研究了我国房地产市场的风险程度及分布情况并做出了评价分析。
[Abstract]:In our country, the real estate industry plays an important role in the national economy, which not only concerns the people's living standard and housing security, but also affects the stability of the economic system in every region and even the whole country in recent years. As a result of various phenomena in the real estate market of our country, the real estate market in our country has some risks that can not be ignored. However, it is relatively difficult to evaluate the degree of risk of real estate market because of its nonlinear characteristics. In this context, this paper aims at the nonlinear characteristics of real estate. A complete and scientific evaluation index system is constructed, and the risk degree of real estate market in our country is evaluated accurately by neural network model. According to the development degree and harmony degree of the real estate market, this paper constructs the evaluation index system of China's real estate market. The BP neural network model is used to model and predict the real estate sample data from 2001 to 2012 in 31 provinces and municipalities of China. With the help of artificial neural network module in Matlab7.0, the establishment of BP neural network model is realized. After testing, it is found that the model has good generalization ability. Finally, the risk of real estate market in 31 provinces and municipalities in China in 2013 is forecasted and analyzed by using the established model. Get the real estate risk degree and distribution of our country: 1) our country real estate market overall risk is bigger, only 5 of 31 provinces and municipalities directly under the Central Government are in the condition of relatively small risk.) the real estate market risk of our country is from the developed area of the east coast. Distributed from large to small to underdeveloped areas in the west. Finally, In view of the situation of real estate market risk in China, the countermeasures and suggestions are put forward from two aspects: economic means and legal means. The main innovation of this paper lies in: according to the development degree and harmony degree of real estate market, a comprehensive and reasonable evaluation index system of real estate market risk is constructed; At the same time, BP neural network model is used to study the risk degree and distribution of real estate market in China from the whole national level.
【学位授予单位】：湘潭大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP183;F299.23

【参考文献】