基于网民关注行为大数据实现对房地产价格的预测研究——以北京市房价为例

Prediction of House Price Based on Internet Users’ Behavioral Big Data——Take Housing Price in Beijing as an Example

网络大数据越来越多地被应用到经济问题的分析中。本文利用网民关注行为的大数据结合机器学习中的神经网络模型对北京市房价走势进行预测和分析。研究表明,利用网络搜索大数据的及时性、高频率优势和神经网络模型拟合复杂变量间关系的能力,能实现对房价走势的高精度预测,并且能够极大地提高预测房价走势中"拐点"的成功率。

More and more web search big data is being applied to the analysis of economic problems.This paper combines the web big data of netizens and the neural network model in machine learning to predict and analyze the trend of Beijing house prices.The result shows that combining timeliness and high granularity of online search big data and the ability of neural network models to fit complex relationships between variables can achieve high-precision prediction of housing price trends and capture the inflection point in house price fluctuations.