基于改进狮群算法和BP神经网络模型的房价预测  

Housing price prediction based on improved lion sw arm algorithm and BP neural netw ork model

作　　者：丁飞 江铭炎[1] DING Fei;JIANG Mingyan(School of Information Science and Engineering,Shandong University,Qingdao 266237,Shandong,China)

机构地区：[1]山东大学信息科学与工程学院,山东青岛266237

出　　处：《山东大学学报（工学版）》2021年第4期8-16,共9页Journal of Shandong University（Engineering Science）

基　　金：国家自然科学基金资助项目(61771293);山东省自然科学基金资助项目(ZR2020MF153);山东省重点资助项目(2019JZZY010111)。

摘　　要：将狮群算法(lion swarm optimization, LSO)与海鸥算法的迁徙机制和螺旋搜索机制结合,增强算法的局部搜索能力;同时增加监督机制,提高算法的全局搜索能力。与粒子群算法和狮群算法对比,在常用的测试函数上验证改进狮群算法的优越性。采用改进后的狮群算法优化BP神经网络模型,对房屋价格预测问题进行研究,通过房屋的户型、面积等相关指标有效地对青岛市的二手房价格进行预测。改进后的狮群算法对BP神经网络的权值和偏置进行优化,提高BP神经网络的收敛速度和训练精度。试验结果表明,提出的螺旋搜索狮群和BP结合算法(spiral search lion swarm optimization-BP, SLSO-BP)模型在房价预测问题上预测效果更好。The lion swarm optimization algorithm combined the migration mechanism and spiral search mechanism of seagull algorithm to enhance the local search ability;the global search performance of lion swarm optimization algorithm was enhanced by adding supervision mechanism. The particle swarm optimization algorithm and the lion swarm optimization algorithm were used as the comparison algorithm, and the advantages of the improved algorithm were verified on the common test functions. The improved lion swarm optimization algorithm was used to optimize the BP neural network model to study the problem of housing price prediction, and the price of second-hand housing in Qingdao could be effectively predicted through relevant indicators such as house type and area. The improved lion swarm optimization algorithm was used to optimize the weights and biases of the BP neural network to improve the convergence speed and training accuracy of the BP neural network. The test results showed that the SLSO-BP model proposed in the study had a better prediction effect on the problem of housing price prediction.

关 键 词：狮群算法 螺旋搜索 监督机制 BP神经网络 房价预测 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]