基于稀疏神经网络的广州市二手楼价影响因素分析  

Analysis on the Influencing Factors of Second-Hand House Price in Guangzhou Based on Sparse Neural Network

作　　者：陆晓炘 Lu Xiaoxin(Guangxi Normal University,Guilin 541000)

机构地区：[1]广西师范大学,桂林541000

出　　处：《现代计算机》2022年第8期68-71,81,共5页Modern Computer

摘　　要：针对传统的二手楼价影响因素分析方法主要是建立线性模型而忽略模型为非线性模型的可能性,而且没有考虑在高维情形下部分因素对二手楼价的影响很小导致模型过参数化等问题,本文结合爬虫和高德地图API获取包括微观因素与宏观因素的广州市二手楼信息,对数据进行预处理,建立稀疏神经网络模型。在选定正则化参数后,对数据进行20次建模,在给定阈值的情况下得出500米内地铁数量,1000米内中小学数量,房屋朝向,有无电梯为其主要影响因素的结论。In view of the traditional second-hand house prices factors analysis method is mainly to establish the possibility of linear model and ignores the model for the nonlinear model,and does not take into account in higher dimensional case little impact of some factors on the second-hand housing prices lead to problems such as parameterized model,combined with the crawler and Scott maps API access including guangzhou,the micro and macro factors of secondhand information,The sparse neural network model is established by pre-processing the data.After the regularization parameters are selected,the data are modeled for 20 times,and the number of subways within 500 meters,the number of primary and secondary schools within 1000 meters,the orienta⁃tion of houses and whether there are elevators are the main influencing factors under the given threshold value.

关 键 词：稀疏神经网络 二手房价 变量选择 

分 类 号：F299.23[经济管理—国民经济]