Table 9 Performance comparisons of different classifiers for the prediction of five types of fishing vessels on the independent testing dataset.

From: Attention-enhanced and integrated deep learning approach for fishing vessel classification based on multiple features

Methods

Accuracy

Precision

Recall

F1_score

SVM

0.7410

0.7562

0.7410

0.7408

RF

0.9020

0.9119

0.9020

0.9034

XGBoost

0.8830

0.8865

0.8830

0.8838

CNN

0.8160

0.8287

0.8160

0.8179

BiLSTM

0.8950

0.9041

0.8950

0.8970

BiGRU

0.8970

0.9028

0.8970

0.8983

BiLSTM-CNN-attention

0.9190

0.9222

0.9190

0.9196

Back to article page