基于人工神经网络的房地产市场预警模型研究——以成都市为例

Research on Early Warning Model of Real Estate Market Based on Artificial Neural Network——Taking Chengdu as an Example

随着社会的不断进步,房地产业迅速繁荣,对房地产市场进行有效的调控显得尤为重要。结合成都市房地产市场的特点,以及人工神经网络容错性、易操作等特性,利用人工神经网络建立房地产市场预警模型。选取成都市房地产业2000~2014年的数据,建立BP人工神经网络模型,将2000~2013年的数据作为训练样本,2014年的数据作为检测样本。通过5000次训练,在第263次时得到最佳结果,实现了误差小于1e-6的训练目标,其仿真效果具有较高的可信度。该BP人工神经网络模型在有效进行市场预测、促进房地产市场可持续发展等方面有较高的实际意义。

With the continuous progress of the society,the real estate industry is booming,so effectively controlling the real estate market is particularly important. According to the characteristics of the real estate market in Chengdu,this study chooses the artificial network to establish the real estate market early warning model to utilize its fault tolerance. Selecting 15 data of real estate industry from the year 2000 to 2014 in Chengdu,the BP artificial neural network model is established,among which the data in the range from 2000 to 2013 were used as the training samples and the data of the 2014 as tested samples. By training 5000 times,got the best result when in the 263 th time,and realized the training goal that error is less than 1e-6,and the simulation result has high credibility. The BP artificial neural network model has a high practical significance in foresting the market and promoting the sustainable development of the real estate market.