摘要
使用传统的BP神经网络进行预测容易发生收敛速度慢、预测精度低、陷入局部最优的可能。对此,阐述了BP神经网络的基本原理,介绍了遗传算法的实现过程,并根据遗传算法的全局搜索能力,优化调整了BP神经网络的初始权值和阈值,分别对传统BP神经网络和改进后的GA-BP神经网络建立了房价预测模型。选取了中国房价及其主要影响因素作为实验数据进行仿真训练,对比了模型的预测效果。实验结果表明,经过遗传算法改进的BP神经网络较传统BP神经网络具有预测精度高、收敛速度快的优点,同时避免了陷入局部最优的缺陷。
Usingtraditional BP neural network for prediction is prone to be slow convergence,low prediction accuracy and easy to fall into local optimum. For this we describe the basic principles of BP neural network,introduce the implementation of genetic algorithm,and adjust the initial weights and thresholds of BP neural network according to the global search ability of genetic algorithm. Respectively,the traditional BP neural network and the improved GA-BP neural network are used to establish the housing price prediction model. Finally,the housing price and its main influencing factors are selected as the experimental data for simulation training,and the prediction effect of the model is compared. The experiment shows that the improved BP neural network has higher prediction accuracy and faster convergence speed than the traditional BP neural network,and avoids the defects of falling into the local optimum.
作者
李春生
李霄野
张可佳
LI Chun-sheng;LI Xiao-ye;ZHANG Ke-jia(School of Computer and Information Technology,Northeast Petroleum University,Daqing 163318,China)
出处
《计算机技术与发展》
2018年第8期144-147,151,共5页
Computer Technology and Development
基金
黑龙江省自然科学基金面上项目(F2015020)
省教育科研规划重点课题(GJB1215013)
关键词
BP神经网络
遗传算法
价格预测
误差分析
BP neural network
genetic algorithm
price forecasting
error analysis