• Laser & Optoelectronics Progress
  • Vol. 59, Issue 2, 0210018 (2022)
Zhigang Su1、2、* and Xuemeng Wang1
Author Affiliations
  • 1College of Electronic Information and Automation, Civil Aviation University of China, Tianjin 300300, China
  • 2Sino-European Institute of Aviation Engineering, Civil Aviation University of China, Tianjin 300300, China
  • show less
    DOI: 10.3788/LOP202259.0210018 Cite this Article Set citation alerts
    Zhigang Su, Xuemeng Wang. Aerial Target Classification Algorithm Based on Double-Layer Feature Selection[J]. Laser & Optoelectronics Progress, 2022, 59(2): 0210018 Copy Citation Text show less
    Framework of air target feature selection and classification algorithm
    Fig. 1. Framework of air target feature selection and classification algorithm
    Structure of ELM
    Fig. 2. Structure of ELM
    Relationship between number of experimental repetitions and average score
    Fig. 3. Relationship between number of experimental repetitions and average score
    Comparison of experimental effects of different transfer functions
    Fig. 4. Comparison of experimental effects of different transfer functions
    Relationship between training error, test error and number of hidden layer neurons
    Fig. 5. Relationship between training error, test error and number of hidden layer neurons
    Comparison of feature quantity and precision of each algorithm
    Fig. 6. Comparison of feature quantity and precision of each algorithm
    Comparison of training time
    Fig. 7. Comparison of training time
    AlgorithmPrecisionStandard deviation
    K-S test+De-relief99.78±0.02
    FCBF97.75±0.08

    EDA

    Lasso

    99.11

    95.10

    ±0.12

    ±0.06

    Table 1. Comparison of accuracy and standard deviation of different feature selection algorithms
    AlgorithmPrecision /%TPRFPR
    Ret-OSELM99.70.9960.016
    LSTM98.50.9830.036
    DNN97.80.9790.055
    Naive Bayes92.10.9350.081
    Decision tree95.60.9620.058
    Table 2. Comparison of classification performance of various classification algorithms
    AlgorithmTest time /ms
    TSFRESH extract features52.13
    TSFRESH extract+K-S test15.87

    TSFRESH extracts+K-S test+relief

    TSFRESH extracts+K-S test+De-relief

    2.11

    1.26

    Table 3. Comparison of test time of different feature extraction algorithms
    Zhigang Su, Xuemeng Wang. Aerial Target Classification Algorithm Based on Double-Layer Feature Selection[J]. Laser & Optoelectronics Progress, 2022, 59(2): 0210018
    Download Citation