基于流形正则化的批量分层编码极限学习机

doi:10.3969/j.issn.1000-1158.2021.07.16

摘要
图/表
参考文献(16)
相关文章 (15)

全文: PDF (735 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要针对极限学习机在处理高维数据时存在内存能耗大、分类准确率低、泛化性差等问题,提出了一种批量分层编码极限学习机算法。首先通过对数据集分批处理,以减小数据维度,降低输入复杂性;然后采用多层自动编码器结构对各批次数据进行无监督编码,以实现深层特征提取;最后利用流形正则化思想构建含有继承因子的流形分类器,以保持数据的完整性,提高算法的泛化性能。实验结果表明,该方法实现简单,在NORB,MNIST和USPS数据集上的分类准确率分别可以达到92.16%、99.35%和98.86%,与其它极限学习机算法对比,在降低计算复杂度和减少CPU内存消耗上具有较明显的优势。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	刘彬
	杨有恒
	刘静
	王卫涛
	刘浩然
	闻岩

关键词 ：计量学, 极限学习机, 高维数据, 批次学习, 无监督编码, 流形正则化

Abstract：Aiming at the problems of high memory energy consumption,low classification accuracy and poor generalization when dealing with high-dimensional data, a batch hierarchical coding extreme learning machine algorithm is proposed.Firstly,the dataset is processed in batches to reduce the data dimension and reduce input complexity.Then,the multi-layer automatic encoder structure is used to unsupervise the batch data to achieve deep feature extraction.Finally,the manifold regularization is used to construct a manifold classifier with inheritance factors to maintain data integrity and improve the generalization performance of the algorithm.The experimental results show that the method is simple to implement,and the classification accuracy on the NORB,MNIST,and USPS datasets can reach 92.16%,99.35%,and 98.86%,respectively.Compared with other ELM algorithms,it has obvious advantages in reducing computational complexity and reducing CPU memory consumption.

Key words： metrology extreme learning machine high dimension data batch learning unsupervised coding;manifold regularization

收稿日期: 2019-11-25 发布日期: 2021-07-15

PACS:

TB973

基金资助:河北省自然科学基金(F2019203320,E2018203398);河北省人才培养项目(A201903005)

通讯作者: 杨有恒(1996-),男,陕西汉中人,燕山大学在读硕士研究生,主要研究方向为机器学习,人工智能优化算法。Email: yangyouheng@stumail.ysu.edu.cn E-mail: yangyouheng@stumail.ysu.edu.cn

作者简介: 刘彬(1953-),男,黑龙江哈尔滨人,燕山大学教授,主要从事神经网络算法,数据挖掘方面的研究。Email: liubin311@163.com

引用本文:

刘彬,杨有恒,刘静,王卫涛,刘浩然,闻岩. 基于流形正则化的批量分层编码极限学习机[J]. 计量学报, 2021, 42(7): 937-943.
LIU Bin,YANG You-heng,LIU Jing,WANG Wei-tao,LIU Hao-ran,WEN Yan. Batch Hierarchical Coding Extreme Learning Machine Based on Manifold Regularization. Acta Metrologica Sinica, 2021, 42(7): 937-943.

链接本文:

http://jlxb.china-csm.org:81/Jwk_jlxb/CN/10.3969/j.issn.1000-1158.2021.07.16 或 http://jlxb.china-csm.org:81/Jwk_jlxb/CN/Y2021/V42/I7/937

［1］Huang G B, Zhu Q Y, Siew C K. Extreme learning machine: Theory and applications ［J］. Neurocomputing, 2006, 70(1): 489-501.
［2］Iosifidis A, Tefas A, Pitas I. Graph embedded extreme learning machine ［J］. IEEE Transactions on Cybernetics, 2017, 46 (1): 311-324.
［3］何群, 王红, 江国乾, 等. 基于相关主成分分析和极限学习机的风电机组主轴承状态监测研究［J］. 计量学报, 2018, 39 (1): 89-93.
He Q, Wang H, Jiang G Q, et al. Research of wind turbine main bearing condition monitoring based on correlation PCA and ELM ［J］. Acta Metrologica Sinica, 2018, 39 (1): 89-93.
［4］吴忠强, 尚梦瑶, 申丹丹, 等. 基于BSA-RELM的纯电动汽车锂离子电池SOC估计［J］. 计量学报, 2019, 40 (4): 693-699.
Wu Z Q, Shang M Y, Shen D D, et al. Estimation of SOC of Li-ion battery in pure electric vehicle by BSA-RELM ［J］. Acta Metrologica Sinica, 2019, 40 (4): 693-699.
［5］牛培峰, 丁翔, 刘楠, 等. 基于混合鸡群算法和核极端学习机的锅炉NOx排放的预测［J］.计量学报, 2019, 40 (5): 929-936.
Niu P F, Ding X, Liu N, et al. Prediction of boiler NOx emission based on mixed chicken swarm algorithm and kernel extreme learning machines ［J］. Acta Metrologica Sinica, 2019, 40 (5): 929-936.
［6］Kasun L L C, Zhou H M, Huang G B, et al. Represen-tational learning with extreme learning machine for big data ［J］. IEEE Intelligent Systems, 2014, 28 (6): 30-59.
［7］Tang J X, Deng C W, Huang G B. Extreme learning machine for multilayer perceptron ［J］. IEEE Transac-tions on Neural Networks and Learning Systems, 2016, 27 (4): 809-821.
［8］徐睿, 梁循, 齐金山, 等. 极限学习机前沿进展与趋势［J］. 计算机学报, 2019, 42 (7): 1640-1670.
Xu R, Liang X, Qi J S, et al. Advances and trends in extreme learning machine ［J］.Chinese Journal of Compu-ters, 2019, 42 (7): 1640-1670.
［9］Zhou H, Huang G B, Lin Z, et al. Stacked extreme learning machines ［J］. IEEE Transactions on Cybernetics, 2015, 45 (9): 2013-2025.
［10］Vong C M, Chen C, Wong P K . Empirical kernel map-based multilayer extreme learning machines for represen-tation learning ［J］. Neurocomputing, 2018, 310: 265-276.
［11］Tang G J, Wang X L, He Y L. A novel method of fault diagnosis for rolling bearing based on dual tree complex wavelet packet transform and improved multiscale per-mutation entropy ［J］. Mathematical Problems in Engine-ering, 2016: 1-13.
［12］Huang G B, Zhou H M, DING X J, et al. Extreme learning machine for regression and multiclass classifica-tion ［J］. IEEE Transactions on Systems, Man and Cyb-ernetics, Part B (Cybernetics), 2012, 42 (2): 513-529.
［13］Beck A, Teboulle M. A fast iterative shrinkagethresh-olding algorithm for linear inverse problems ［J］. SIAM Journal on Imaging Sciences, 2009, 2 (1): 183-202.
［14］Yao L, Ge Z . Deep learning of semi-supervised process data with hierarchical extreme learning machine and soft sensor application ［J］. IEEE Transactions on Industrial Electronics, 2017, 65 (2): 1490-1498.
［15］Huang G, Song S, Gupta J N D, et al. Semi-supervised and unsupervised extreme learning machines ［J］. IEEE Transactions on Cybernetics, 2014, 44 (12): 2405-2417.
［16］张立峰,朱炎峰. 基于粒子群优化极限学习机及电容层析成像的两相流流型及其参数预测［J］.计量学报, 2020, 41 (12): 1488-1493.
Zhang L F, Zhu Y F. Two-phase Flow Regime and its Parameter Prediction Based on Particle Swarm Optimization Extreme Learning Machine and Electrical Capacitance Tomography［J］. Acta Metrologica Sinica, 2020, 41 (12): 1488-1493.