马尔可夫链在地价预测模型中的应用研究

在地价的预警预报研究过程中,过去的理论主要存在3点不足,（1）往往没有建立科学完善的数学模型,也没有用多种数学模型相互验证预测结论;（2）没有充分利用现有的信息,例如地价监测体系的信息.（3）在实际应用中也存在预测地价水平与城市实际地价水平、预测走势与实际变化有偏差的问题.解决这些问题通过以下方法：充分利用现有的准确的地价监测体系的信息,并在此基础上建立时间序列模型、空间分布预测模型及马尔可夫链地价预测模型;以不动产周期理论和专家经验知识库检验经过数学模型测算的结论.马尔可夫链地价预测模型是概率分布式的预测模型,它专门用于分析系统中不确定性因素的概率分布,经过检验可以认为这个模型适用于对地价变化进行科学的概率预测.

Research of Markov Chain Application in Land Price Warning Model

There are two main disadvantages during the process of researching the land price warning system.Firstly,the integrated mathematical model hasn't been set up during the construction of the system,and the real estate cycle theory hasn't been introduced.Secondly,the current information hasn't been well utilized,such as the information about dynamic monitoring system of land price.When it is applied,there still exists the deviation between the forecasted land value and the actual land value,as well as between the forecasted trend and the real change.This paper gives some solutions as follows:(1) make full use of the existing accurate information of land price dynamic monitoring system,and set up time-series models and forecast models of spatial distribution.(2) verify the conclusions by the real estate cycle theory and expertise database.Compared with some traditional time-series forecasting methods,Markov chain models are better at analyzing and forecasting the volatility and trend of land price,which provide an efficient approach for us to comprehend the market movements.To apply the theory of Markov chain to establishing mathematical model of land price's fluctuation and anticipation,we propose the theory of the cycle of land-price's motion and maximization of administrator's and our government's gain. Examples proves that this model and its application are feasible.