城市轨道交通沿线住宅房产价格预测模型研究

发布时间：2018-02-01 14:35

本文关键词： 城市轨道交通住宅房产价格预测 BP神经网络马尔可夫链　出处：《北京交通大学》2014年硕士论文　论文类型：学位论文

【摘要】：我国正处于城市化加速发展阶段和社会经济的转型时期,经济的快速增长加快了城市交通的发展速度。城市轨道交通已经成为我国解决城市交通问题的长期发展战略,在各大城市掀起了建设浪潮。城市轨道交通对沿线土地利用性质的影响使得其沿线范围服务影响区域内住宅、商业用地等的需求量猛增,提高了土地开发强度。这些改变使得城市轨道交通沿线房地产的增值潜力大大提高。目前,大多数针对城市轨道交通沿线房地产价格的定量研究采用的基本都是交通成本模型和特征价格模型。然而,由于模型建立所需的大量基础数据可得性较差,使用这些模型研究我国城市轨道交通沿线影响范围内的房地产价格具有一定局限性。因此,在基础数据有限的情况下,本文结合多元统计分析方法,使用BP神经网络和马尔可夫链建立了城市轨道交通沿线住宅房产价格预测模型。本文通过研究城市轨道交通和土地利用之间的关系,结合大量研究学者的研究成果,分析了影响城市轨道交通沿线住宅房产价格的因素,认为影响城市轨道交通沿线住宅房产价格的因素主要包括区位因素、结构因素和邻里因素,并从这三类因素中选取了城市轨道交通沿线住宅房产价格的影响指标,并对影响指标进行了因子分析,将其简化为6个共性因子作为BP神经网络的输入层神经元。随后,本文利用BP神经网络的理论构造了城市轨道交通沿线住宅房产价格预测模型的BP神经网络结构,并建立了基于马尔可夫链的修正模型,以提高预测精度。最后,本文运用建立的土地资源价值预测模型对10处住宅房产的样本数据进行了分析,预测了这些样本的土地价值。经过实际价值与模型的预测结果比对,验证了该模型能够有效的预测城市轨道沿线住宅的价格,从而为解决URRT规划开发部门部分资金短缺、市民盲目选购房产、跟风投资,房地产商缺乏对房地产开发项目获利能力了解等问题提供有效的科学依据。图8幅,表17个,参考文献54篇。
[Abstract]:China is in the period of accelerating urbanization and social and economic transformation. The rapid growth of economy has accelerated the development of urban transportation. Urban rail transit has become a long-term development strategy to solve urban traffic problems in China. The impact of urban rail transit on the nature of land use along the route makes the demand for residential and commercial land increase rapidly. These changes have greatly increased the value added potential of real estate along urban rail transit. At present, most of the quantitative research on real estate prices along the urban rail transit line is based on the transport cost model and characteristic price model. Because of the poor availability of a large number of basic data needed to establish the model, it is limited to use these models to study the real estate price along the influence range of urban rail transit in China. In the case of limited basic data, this paper uses BP neural network and Markov chain to establish the housing property price prediction model along the urban rail transit line combined with the multivariate statistical analysis method. Through the study of the relationship between urban rail transit and land use, combined with the research results of a large number of scholars, this paper analyzes the factors that affect the housing prices along the urban rail transit line. It is considered that the main factors influencing the residential property prices along the urban rail transit line include location factors, structural factors and neighborhood factors. And from these three factors selected the urban rail transit along the residential property price impact indicators and factors analysis of the impact indicators. It is simplified to six common factors as input layer neurons of BP neural network. In this paper, the BP neural network structure of residential property price prediction model along urban rail transit is constructed by using the theory of BP neural network, and the modified model based on Markov chain is established to improve the prediction accuracy. Finally, this paper uses the land resource value prediction model to analyze the sample data of 10 residential properties, and forecasts the land value of these samples. The actual value is compared with the forecast results of the model. It is verified that the model can effectively predict the housing prices along the urban track, so as to solve the URRT planning and development department part of the shortage of funds, citizens blindly buy real estate, follow the trend of investment. Real estate developers lack effective scientific evidence to understand the profitability of real estate development projects. There are 8 figures, 17 tables and 54 references.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F299.23

【参考文献】