基于BP神经网络的商品房销售量预测研究  

Study on Prediction of Commercial Housing Sale Volume Based on BP Neural Network

作　　者：陈莉[1] 李斌[1] 李广[1] 

机构地区：[1]华中科技大学土木工程与力学学院,武汉430074

出　　处：《基建优化》2005年第4期56-58,共3页Optimization of Capital Construction

摘　　要：商品房销售量受多个因素的作用,而且这些作用多是非线性的。商品房销售量预测实质上是实现一个非线性的映射。本文选择了恰当的经济指标作为BP网络的输入输出变量,探讨将神经网络技术应用于商品房销售量的预测领域,并与回归模型、Logistic模型的预测结果进行比较,结果表明人工神经网络方法在商品房销售量预测中的应用是可行的,预测效果更好。Commercial housing sale volume is affected by several nonlinear factors. The prediction ot the commercial nousmg sale volume is essentially to realize a nonlinear mapping from the input space to the output space. This paper use some appropriate economic indicators as input and output of BP neural network , approach the application of neural network in the prediction of the commercial housing sale volume, the forecasting result is more effective than traditional prediction methods, such as Linear regressive Model, Logistic Model, etc.

关 键 词：商品房销售量 预测 BP神经网络 

分 类 号：F293.3[经济管理—国民经济]