|本期目录/Table of Contents|

[1]李 蓓,张兴敢,方 晖*. 一种改进的BP-Adaboost算法及在雷达多目标分类上的应用[J].南京大学学报(自然科学),2017,53(5):984.[doi:10.13232/j.cnki.jnju.2017.05.018]
 Li Bei,Zhang Xinggan,Fang Hui*. An improved algorithm of BP-Adaboost and application of radar multi-target classification[J].Journal of Nanjing University(Natural Sciences),2017,53(5):984.[doi:10.13232/j.cnki.jnju.2017.05.018]
点击复制

 一种改进的BP-Adaboost算法及在雷达多目标分类上的应用()
     

《南京大学学报(自然科学)》[ISSN:0469-5097/CN:32-1169/N]

卷:
53
期数:
2017年第5期
页码:
984
栏目:
出版日期:
2017-09-30

文章信息/Info

Title:
 An improved algorithm of BP-Adaboost and application of radar multi-target classification
作者:
 李 蓓张兴敢方 晖*
 南京大学电子科学与工程学院,南京,210093
Author(s):
 Li BeiZhang XingganFang Hui*
 School of Electronic Science and Engineering,Nanjing University,Nanjing,2100093,China
关键词:
 Adaboost雷达高分辨率距离像多分类BP神经网络
Keywords:
 Adaboosthigh resolution range profile(HRRP)multi-classBP neural networks
分类号:
TN957.52
DOI:
10.13232/j.cnki.jnju.2017.05.018
文献标志码:
A
摘要:
 基于BP-Adaboost的目标分类算法用于雷达目标分类具有良好的效果.随着训练样本以及测试样本数增加,经典“一对多(One vs.Rest,OvR)”BP-Adaboost算法所需用时也随之增加.提出一种改进的多分类BP-Adaboost算法应用在雷达多目标分类上,在提高分类准确率的同时,有效地解决经典算法在多分类上时间开销过大的问题.该方法采用二进制方法重新表示样本数据类别,使用Adaboost算法将多个BP神经网络弱分类器集成起来学习,通过修改经典算法中的损失函数连续调整训练样本分布和弱分类器的权重,最终形成一个强分类器.对雷达高分辨率距离像(High Resolution Range Profile,HRRP)数据集进行分类仿真结果表明,相比于单个BP神经网络基学习器,所提算法的分类准确率提高了5%~10%,相比于经典的“一对多”BP-Adaboost算法,该算法所需用时仅为传统算法的1/2~1/3.
Abstract:
 BP-Adaboost based classification algorithm has a promising performance in the field of Radar target classification.However,due to the increase of the volume of training samples and testing samples,traditional BP-Adaboost based algorithm OvR(One vs.Rest)requires an amount of time.Therefore,this paper proposes an improved BP-Adaboost algorithm to solve the issue of Radar multi-target classification,which can improve accuracy and solve the time issue that traditional algorithm will encounter in multi-class classification effectively.Specially,we preprocess the initial categories in binary,then several BP neural networks are boosted up to form a new strong classifier,in which process we keep adjusting the distribution of training samples and weight of weak classifiers by modifying the loss function and the number of output nodes.Experiments using the range profile datasets have been demonstrated that the proposed scheme can raise accuracy up to 5%~10% compared to single BP neural network,and significantly improves the time issue efficiently compared to traditional algorithm OvR,which can save 1/2~2/3 running time under the same condition.

参考文献/References:

 [1] Skolnik M I.Introductin to radar systems.McGraw-Hill Electrical Engineering Series,2001(2):38-44.
[2] 刘记红,徐少坤,高勋章等.压缩感知雷达成像技术综述.信号处理,2011,27(2):251-260.(Liu J H,Xu S K,Gao X Z,et al.A review of radar imaging technique based on compressed sensing.Signal Processing,2011,27(2):251-260.)
[3] Du L,Liu H W,Bao Z,et al.Radar HRRP target recognition based on higher order spectra.IEEE Transactions on Signal Processing,2005,53(7):2359-2368.
[4] Xing M D,Bao Z,Pei B N.Properties of high-resolution range profiles.Optical Engineering,2002,41(2):493-504.
[5] Breiman L,Friedman J,Stone C J,et al.Classification and regression trees.Belmont,CA,USA:CRC Press,1984,368.
[6] Bishop C M.Neural networks for pattern recognition.Oxford,UK:Oxford University Press,1996,504.
[7] Cortes C,Vapnik V.Support-vector networks.Machine Learning,1995,20(3):273-297.
[8] Bishop C M.Pattern recognition:Machine learning.Springer New York,2006:1-58.
[9] Wang X D,Wu C M,Qi Y,et al.HRRP classification by using improved SVM decision tree.In:Proceedings of the 6th World Congress on Intelligent Control and Automation.Dalian,China:IEEE,2006:10096-10100.
[10] 陈 凤,杜 兰,保 铮.一种优化K近邻准则及在雷达HRRP目标识别中的应用.西安电子科技大学学报,2007,34(5):681-686.(Chen F,Du L,Bao Z.Modified KNN rule with its application in radar HRRP target recognition.Journal of Xidian University,2007,34(5):681-686.)
[11] Demiriz A,Bennett K P,Shawe-Taylor J.Linear programming boosting via column generation.Machine Learning,2002,46(1-3):225-254.
[12] Freund Y,Schapire R E.A Decision-theoretic generalization of on-line learning and an application to boosting.Journal of Computer and System Sciences,1997,55(1):119-139.
[13] Bengio Y,Courville A,Vincent P.Representation learning:A review and new perspectives.IEEE Transactions on Pattern Analysis and Machine Intelligence,2013,35(8):1798-1828.
[14] Lü Y M,Tang D Z,Xu H,et al.Productivity matching and quantitative prediction of coalbed methane wells based on BP neural network.Science China Technological Sciences,2011,54(5):1281-1286.
[15] Yu J Y.The application of BP-Adaboost strong classifier to acquire knowledge of student creativity.In:Proceedings of 2011 International Conference on Computer Science and Service System(CSSS).Nanjing,China:IEEE,2011:2669-2672.
[16] 曹 莹,苗启广,刘家辰等.AdaBoost算法研究进展与展望.自动化学报,2013,39(6):745-758.(Cao Y,Miao Q G,Liu J C,et al.Advance and prospects of AdaBoost algorithm.Acta Automatica Sinica,2013,39(6):745-758.)
[17] 李 翔,朱全银.Adaboost算法改进BP神经网络预测研究.计算机工程与科学,2013,35(8):96-102.(Li X,Zhu Q Y.Prediction of improved BP neural network by Adaboost algorithm.Computer Engineering & Science,2013,35(8):96-102.)

相似文献/References:

备注/Memo

备注/Memo:
 基金项目:江苏省基础研究计划(BK20151391)
收稿日期:2017-06-25
*通讯联系人,E-mail:fanghui@nju.edu.cn
更新日期/Last Update: 2017-09-25