• Journal of Atmospheric and Environmental Optics
  • Vol. 19, Issue 1, 62 (2024)
DUAN Peijie1、2, LI Zerui2、*, LI Kun3, XU Zhenyi2, LYU Zhao4, and KANG Yu2、3
Author Affiliations
  • 1School of Artificial Intelligence, Anhui University, Hefei 230601, China
  • 2Institute of Artificial Intelligence, Hefei Comprehensive National Science Center, Hefei 230088, China
  • 3Institute of Advanced Technology, University of Science and Technology of China, Hefei 230088, China
  • 4School of Computer Science and Technology, Anhui University, Hefei 230601, China
  • show less
    DOI: 10.3969/j.issn.1673-6141.2024.01.005 Cite this Article
    Peijie DUAN, Zerui LI, Kun LI, Zhenyi XU, Zhao LYU, Yu KANG. On-road high-emitter identification method based on mixed kernel extreme learning machine[J]. Journal of Atmospheric and Environmental Optics, 2024, 19(1): 62 Copy Citation Text show less
    Diagram of on-road remote sensing detection equipment
    Fig. 1. Diagram of on-road remote sensing detection equipment
    Diagram of ELM network structure
    Fig. 2. Diagram of ELM network structure
    Visualization diagram of data distribution after dimensionality reduction using t-SNE
    Fig. 3. Visualization diagram of data distribution after dimensionality reduction using t-SNE
    Diagram of KELM decision region with four different kernel functions. (a) Polynomial kernel function; (b) Sigmoid kernel function; (c) Gaussian kernel function; (d) Laplacian kernel function
    Fig. 4. Diagram of KELM decision region with four different kernel functions. (a) Polynomial kernel function; (b) Sigmoid kernel function; (c) Gaussian kernel function; (d) Laplacian kernel function
    Diagram of MKELM decision region with four different mixed kernel functions. (a) Gaussian-Gaussian mixed kernel function; (b) Laplacian-Laplacian mixed kernel function; (c) Gaussian-Laplacian mixed kernel function
    Fig. 5. Diagram of MKELM decision region with four different mixed kernel functions. (a) Gaussian-Gaussian mixed kernel function; (b) Laplacian-Laplacian mixed kernel function; (c) Gaussian-Laplacian mixed kernel function
    MethodSF1/%RFA/%RMA/%
    SVM79.40 ± 2.2417.87 ± 4.7622.44 ± 7.98
    RF78.48 ± 3.8315.88 ± 7.5925.94 ± 5.45
    ELM58.50 ± 4.8240.00 ± 11.0741.44 ± 4.58
    KELM (G)79.86 ± 4.7320.00 ± 8.3819.53 ± 7.15
    KELM (L)79.74 ± 4.5319.78 ± 8.3320.01 ± 6.81
    MKELM (G+G)81.10 ± 1.7616.48 ± 4.9220.78 ± 4.85
    MKELM (L+L)80.59 ± 3.3515.86 ± 6.7722.20 ± 5.46
    MKELM (G+L)81.40 ± 2.5615.50 ± 5.2221.10 ± 5.14
    Table 1. Comparison of identification results of eight methods
    Peijie DUAN, Zerui LI, Kun LI, Zhenyi XU, Zhao LYU, Yu KANG. On-road high-emitter identification method based on mixed kernel extreme learning machine[J]. Journal of Atmospheric and Environmental Optics, 2024, 19(1): 62
    Download Citation