A novel active learning approach for the classification of hyperspectral imagery using quasi-Newton multinomial logistic regression

被引:13
|
作者
Tan, Kun [1 ]
Wang, Xue [1 ,2 ]
Zhu, Jishuai
Hu, Jun [3 ]
Li, Jun [4 ]
机构
[1] China Univ Min & Technol, Key Lab Land Environm & Disaster Monitoring NASG, Xuzhou 221116, Peoples R China
[2] Chang Guang Satellite Technol Co Ltd, Data Ctr, Sect 3, Changchun, Jilin, Peoples R China
[3] NASG, Inst Aero Photogrammetry & Remote Sensing 1, Xian, Shaanxi, Peoples R China
[4] Sun Yat Sen Univ, Sch Geog & Planning, Ctr Integrated Geog Informat Anal, Guangdong Prov Key Lab Urbanizat & Geosimulat, Guangzhou, Guangdong, Peoples R China
关键词
REMOTE-SENSING IMAGES; SPECTRAL-SPATIAL CLASSIFICATION; RANDOM FOREST CLASSIFIER; SUPPORT VECTOR MACHINES; SEMISUPERVISED CLASSIFICATION; DISCRIMINANT-ANALYSIS; EM ALGORITHM; INFORMATION; ACCELERATION; SEGMENTATION;
D O I
10.1080/01431161.2018.1433893
中图分类号
TP7 [遥感技术];
学科分类号
081102 ; 0816 ; 081602 ; 083002 ; 1404 ;
摘要
In this article, a novel active learning approach is proposed for the classification of hyperspectral imagery using quasi-Newton multinomial logistic regression/Davidon, Fletcher, and Powell selective variance (MLR-DFP-SV). The proposed approach consists of two main steps: (1) a fast solution for the MLR classifier, where the logistic regressors are obtained by the use of the quasi-Newton algorithm; and (2) selection of the most informative unlabelled samples. The SV method is applied to select the most informative unlabelled samples, based on the posterior density distributions. Experiments on two real hyperspectral data sets confirmed that the proposed approach can effectively select the most informative unlabelled samples and improve the classification accuracy. Three different methods - the maximum information (MI), breaking ties (BT), and minimum error (ME) methods - were also used to obtain the most informative unlabelled samples, and it was found that the new sample selection method - SV - can select more informative samples than the BT, MI, and ME methods.
引用
收藏
页码:3029 / 3054
页数:26
相关论文
共 50 条
  • [31] A Quasi-Newton method for solving generalized equations by using a Kantorovich approach
    Amaral, V. S.
    Santos, P. S. M.
    Silva, G. N.
    Souza, S. S.
    COMPUTATIONAL & APPLIED MATHEMATICS, 2024, 43 (02):
  • [32] Classification of the Types of Defects in Steam Generator Tubes using the Quasi-Newton Method
    Lee, Joonpyo
    Jo, Nam H.
    Roh, Youngsu
    JOURNAL OF ELECTRICAL ENGINEERING & TECHNOLOGY, 2010, 5 (04) : 666 - 671
  • [33] An active learning approach to hyperspectral data classification
    Rajan, Suju
    Ghosh, Joydeep
    Crawford, Melba M.
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2008, 46 (04): : 1231 - 1242
  • [34] A novel logistic regression model combining semi-supervised learning and active learning for disease classification
    Chai, Hua
    Liang, Yong
    Wang, Sai
    Shen, Hai-wei
    SCIENTIFIC REPORTS, 2018, 8
  • [35] A novel logistic regression model combining semi-supervised learning and active learning for disease classification
    Hua Chai
    Yong Liang
    Sai Wang
    Hai-wei Shen
    Scientific Reports, 8
  • [36] SEMI-SUPERVISED HYPERSPECTRAL IMAGE CLASSIFICATION BASED ON A MARKOV RANDOM FIELD AND SPARSE MULTINOMIAL LOGISTIC REGRESSION
    Li, Jun
    Bioucas-Dias, Jose M.
    Plaza, Antonio
    2009 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM, VOLS 1-5, 2009, : 2119 - +
  • [37] Sugarcane Land Classification with Satellite Imagery using Logistic Regression Model
    Henry, F.
    Herwindiati, D. E.
    Mulyono, S.
    Hendryli, J.
    INTERNATIONAL CONFERENCE ON INFORMATION TECHNOLOGY AND DIGITAL APPLICATIONS, 2017, 185
  • [38] CLASSIFICATION OF PSYCHIATRIC DISORDERS USING MULTINOMIAL LOGISTIC REGRESSION VERSUS ARTIFICIAL NEURAL NETWORK
    Martin Perez, Elena
    Caldero Alonso, Amaya
    Martin Martin, Quintin
    JP JOURNAL OF BIOSTATISTICS, 2021, 18 (03) : 395 - 408
  • [39] A novel framework for classification of two-class motor imagery EEG signals using logistic regression classification algorithm
    Khan, Rabia Avais
    Rashid, Nasir
    Shahzaib, Muhammad
    Malik, Umar Farooq
    Arif, Arshia
    Iqbal, Javaid
    Saleem, Mubasher
    Khan, Umar Shahbaz
    Tiwana, Mohsin
    PLOS ONE, 2023, 18 (09):
  • [40] AN NOVEL ACTIVE LEARNING STRATEGY FOR HYPERSPECTRAL IMAGE CLASSIFICATION
    Shi, Qian
    Zhang, Liangpei
    Du, Bo
    2012 4TH WORKSHOP ON HYPERSPECTRAL IMAGE AND SIGNAL PROCESSING (WHISPERS), 2012,