噪声可容忍的标记组合半监督学习算法

doi:10.19678/j.issn.1000-3428.0050398

计算机工程 ›› 2019, Vol. 45 ›› Issue (4): 157-162,168. doi: 10.19678/j.issn.1000-3428.0050398

噪声可容忍的标记组合半监督学习算法

林金钏,艾浩军

武汉大学计算机学院,武汉 430072

收稿日期:2018-02-02 出版日期:2019-04-15 发布日期:2019-04-15
作者简介:林金钏(1992—),男,硕士研究生,主研方向为迁移学习、复杂网络;艾浩军,副教授。
基金资助:
国家重点研发计划(2016YFB0502201)。

Noise Tolerant Label Combination Semi-supervised Learning Algorithm

LIN Jinchuan,AI Haojun

School of Computer Science,Wuhan University,Wuhan 430072,China

Received:2018-02-02 Online:2019-04-15 Published:2019-04-15

摘要/Abstract

摘要：

针对传统机器学习方法在完成分类任务时多数存在人工标记成本较高、泛化能力较弱的问题,提出一种标记组合半监督学习算法。基于集成学习的思想,利用有标记数据训练多个弱模型并进行组合,增强模型的泛化能力。对无标记数据进行预测,生成有噪声的标记并组合建模。在风险最小化的框架下,使模型收敛达到最优。实验结果表明,在2种有监督场景下与现有的支持向量机、分类与回归树、神经网络等算法相比,该算法具有较优的泛化能力。

关键词: 半监督学习, 集成学习, 风险最小化, 梯度下降, 损失函数

Abstract:

Traditional machine learning method always needs high cost manual marking process,and exhibits weak ability of generalization in classification task.In order to solve these problems,a label combination semi-supervised learning algorithm is proposed.Taking advantage of the principle of ensemble learning,the algorithm uses the labeled data to train multiple weak learners,and combine them to enhance the generalization ability.Predict the unlabeled data to generate noise labels,and then combine and model these noise labels to make the model more robust.Under the framework of risk minimization,the model converges to the optimal state.Experimental results show that,compared with some existing learning algorithms like Support Vector Machine(SVM),Classification and Regression Tree(CART),Neural Network(NN),the algorithm has relatively good generalization ability.

Key words: semi-supervised learning, ensemble learning, risk minimization, gradient descent, loss function

中图分类号:

TP181

林金钏,艾浩军. 噪声可容忍的标记组合半监督学习算法[J]. 计算机工程, 2019, 45(4): 157-162,168.

LIN Jinchuan,AI Haojun. Noise Tolerant Label Combination Semi-supervised Learning Algorithm[J]. Computer Engineering, 2019, 45(4): 157-162,168.

https://www.ecice06.com/CN/Y2019/V45/I4/157

参考文献

［1］张晨光,张燕.半监督学习［M］.北京:中国农业科学技术出版社,2013.
［2］NIGAM K,MCCALLUM A K,THRUN S,et al.Text classification from labeled and unlabeled documents using EM［J］.Machine Learning,2000,38(2/3):103-134.
［3］BLUM A,MITCHELL T.Combining labeled and unlabeled data with co-training［C］//Proceedings of the 11th Conference on Computational Learning Theory.New York,USA:ACM Press,1998:92-100.
［4］JOACHIMS T.Transductive inference for text classifica-tion using support vector machines［C］//Proceedings of International Conference on Machine Learning.San Francisco,USA:Morgan Kaufmann Publishers,1999:200-209.
［5］ZHU X J,GHAHRAMANI Z,LAFFERTY J D.Semi-supervised learning using gaussian fields and harmonic functions［C］//Proceedings of the 12th International Conference on Machine Learning.Palo Alto,USA:AAAI Press,2003:912-919.
［6］ZHOU Z H.Ensemble methods:foundations and algorithms［M］.London,UK:Taylor and Francis Group,2012.
［7］FRNAY B,VERLEYSEN M.Classification in the presence of label noise:a survey［J］.IEEE Transactions on Neural Networks and Learning Systems,2014,25(5):845-869.
［8］刘康,钱旭,王自强.主动学习算法综述［J］.计算机工程与应用,2012,48(34):1-4.
(下转第168页)
(上接第162页)
［9］BREIMAN L.Bagging predictors［J］.Machine Learning,1996,24(2):123-140.
［10］SCHAPIRE R E,FREUND Y,BARTLETT P,et al.Boosting the margin:a new explanation for the effectiveness of voting methods［C］//Proceedings of the 14th International Conference on Machine Learning.San Francisco,USA:Morgan Kaufmann Publishers,1997:322-330.
［11］MANWANI N,SASTRY P S.Noise tolerance under risk minimization［J］.IEEE Transactions on Cybernetics,2013,43(3):1146-1151.
［12］TEWARI A,BARTLETT P L.On the consistency of multiclass classification methods［C］//Proceedings of International Conference on Computational Learning Theory.Berlin,Germany:Springer,2007:143-157.
［13］SHAWE-TAYLOR J,BARTLETT P L,WILLIAMSON R C,et al.Structural risk minimization over data-dependent hierarchies［J］.IEEE Transactions on Information Theory,1998,44(5):1926-1940.
［14］WESTON J,WATKINS C.Multi-class support vector machines［C］//Proceedings of European Symposia on Artificial Neural Networks.Brussels,Belgium:［s.n.］,1999:83-128.
［15］BARTLETT P L,JORDAN M I,MCAULIFFE J D.Convexity,classification,and risk bounds［J］.Journal of the American Statistical Association,2006,101(473):138-156.
［16］SZEGEDY C,IOFFE S,VANHOUCKE V,et al.Inception-v4,inception-resnet and the impact of residual connections on learning［C］//Proceedings of the 31st AAAI Conference on Artificial Intelligence.Palo Alto,USA:AAAI Press,2017:4278-4284.
［17］MALLAPRAGADA P K,JIN R,JAIN A K,et al.SemiBoost:boosting for semi-supervised learning［J］.IEEE Transactions on Pattern Analysis and Machine Intelligence,2009,31(11):2000-2014.
［18］LI Y Y,SU L,CHEN J,et al.Semi-supervised question classification based on ensemble learning［C］//Proceedings of International Conference on Swarm Intelligence.Berlin,Germany:Springer,2015:341-348.
［19］GUYON I.Design of experiments for the NIPS 2003 variable selection benchmark［EB/OL］.［2018-01-05］.http://clopinet.com/isabelle/Projects/NIPS2003/Slides/ NIPS2003-Datasets.pdf.

[1]	屠乃威, 焦猛, 阎馨. 复杂环境下输电线路鸟巢目标图像检测模型[J]. 计算机工程, 2024, 50(7): 216-226.
[2]	顾永跟, 高凌轩, 吴小红, 陶杰. 非独立同分布下联邦半监督学习的数据分享研究[J]. 计算机工程, 2024, 50(6): 188-196.
[3]	杜田田, 王晓龙, 何劲. 复杂光照条件下基于光流的水运航道流速检测算法[J]. 计算机工程, 2024, 50(4): 60-67.
[4]	马明旭, 马宏, 宋华伟. 基于YOLO-Pose的城市街景小目标行人姿态估计算法[J]. 计算机工程, 2024, 50(4): 177-186.
[5]	单永航, 张希, 胡川, 丁涛军, 姚远. 基于集成学习的交通事故严重程度预测研究与应用[J]. 计算机工程, 2024, 50(2): 33-42.
[6]	蒋心璐, 陈天恩, 王聪, 赵春江. 大田环境下的农业害虫图像小目标检测算法[J]. 计算机工程, 2024, 50(1): 232-241.
[7]	沈志东, 岳恒宪. 基于分布式扰动的文本对抗训练方法[J]. 计算机工程, 2023, 49(9): 16-22.
[8]	李嘉新, 侯进, 盛博莹, 周宇航. 基于改进YOLOv5的遥感小目标检测网络[J]. 计算机工程, 2023, 49(9): 256-264.
[9]	包善书, 车波, 邓林红. 基于双源域迁移学习的肺音信号识别[J]. 计算机工程, 2023, 49(9): 295-302, 312.
[10]	陈露萌, 曹彦彦, 黄民, 谢鑫钢. 基于改进YOLOv5的火焰检测方法[J]. 计算机工程, 2023, 49(8): 291-301, 309.
[11]	陈仲磊, 伊鹏, 陈祥, 胡涛. 基于集成学习的系统调用实时异常检测框架[J]. 计算机工程, 2023, 49(6): 162-169,179.
[12]	毕然, 王轶, 周喜. 基于重建误差的任务型对话未知意图检测[J]. 计算机工程, 2023, 49(2): 54-60.
[13]	胡清翔, 饶文碧, 熊盛武. 面向无人机遥感场景的轻量级小目标检测算法[J]. 计算机工程, 2023, 49(12): 169-177.
[14]	陈安琪, 陈睿, 邝祝芳, 黄华军. 基于图神经网络的不平衡欺诈检测研究[J]. 计算机工程, 2023, 49(11): 150-159.
[15]	戚玲珑, 高建瓴. 基于改进YOLOv7的小目标检测[J]. 计算机工程, 2023, 49(1): 41-48.

选择文件类型/文献管理软件名称

选择包含的内容

噪声可容忍的标记组合半监督学习算法

Noise Tolerant Label Combination Semi-supervised Learning Algorithm

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

模态框（Modal）标题

选择文件类型/文献管理软件名称

选择包含的内容

噪声可容忍的标记组合半监督学习算法

Noise Tolerant Label Combination Semi-supervised Learning Algorithm

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价