SNDD: A Sonar Noise Discriminator Based on DeepGBM

In order to realize the automatic classification of sonar waveform and noise interference, a Sonar Noise Discriminator based on DeepGBM (SNDD) is proposed. First, the sonar waveform is characterized by feature extraction; Then, according to the actual needs, SNDD-R, SNDD-P or SNDD-G sub-models are selected to conduct dimensionality reduction processing on the data set; Finally, the data set is entered into DeepGBM for classifier training to complete the classification of the sonar waveform. Validation experiments on the actual sonar waveform data set show that the SNDD model has a better classification effect than other models.