首页 | 期刊介绍 | 编委会 | 道德声明 | 投稿指南 | 常用下载 | 过刊浏览 | In English
引用本文:
【打印本页】   【HTML】   【下载PDF全文】   查看/发表评论  下载PDF阅读器  关闭
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 104次   下载 0  
分享到: 微信 更多
基于决策树与二分分割算法的BP神经网络在赤潮等级预测中的应用研究
李海涛,刘泰麟
青岛科技大学
摘要:
针对赤潮灾害等级预测难的现状,提出了一种基于C4.5决策树与二分分割算法优化的BP(反向传播)神经网络赤潮等级预测模型。该模型针对传统BP神经网络输入参数难以选择和隐含层节点数量难以确定的问题,通过决策树分类获取最优的属性组合,来解决输入参数难以选择的问题;通过“二分分割算法”,来解决隐含层节点数难以确定的问题。实验结果表明,该模型在青岛近海海域赤潮灾害等级预测中,预测结果的均方根误差(RMSE)小于传统BP神经网络的预测误差,并且在网络训练时间上有所缩短,预测精度上有所提高,能够获得良好的预测结果,可为赤潮等级预测提供新的解决方法。
关键词:  赤潮灾害等级预测  C4.5决策树算法  二分分割算法  BP神经网络  
DOI:10.11759/hykx20181130001
分类号:
基金项目:青岛市创新创业领军人才;农业部水产养殖数字农业建设试点项目
Application of BP neural network based on decision tree and two partition algorithm in red tide classification prediction Research
lihaitao1,2, liutailin1,2
1.Qingdao University of Science &2.Technology
Abstract:
Aiming at the difficulty of red tide disaster grade prediction, a BP (back propagation) neural network model based on C4.5 decision tree and dichotomy algorithm is proposed. Aiming at the problem that input parameters of traditional BP neural network are difficult to select and the number of hidden layer nodes is difficult to determine, the model solves the problem that input parameters are difficult to select by using decision tree classification to obtain the optimal combination of attributes, and solves the problem that the number of hidden layer nodes is difficult to determine by using "two partition algorithm". The experimental results show that the root mean square error (RMSE) of the prediction results is less than that of the traditional BP neural network, and the training time of the network is shortened, and the prediction accuracy is improved. Good prediction results can be obtained, which can be used to predict the red tide level. Measurements provide new solutions.
Key words:  Red tide disaster grade prediction  C4.5 decision tree algorithm  two partition algorithm  BP neural network  
版权所有 《海洋科学》编辑部 Copyright©2008 All Rights Reserved
主管单位:中国科学院 主办单位:中国科学院海洋研究所
地址:青岛市南海路七号  邮编:266071  电话:0532-82898755  E-mail:marinesciences@qdio.ac.cn
技术支持:北京勤云科技发展有限公司