Table 13 The accuracy result with and without imbalance learning (%)
From: Exploring best-matched embedding model and classifier for charging-pile fault diagnosis
Classifiers* | GloVe | Word2vec | BERT | |
|---|---|---|---|---|
RF | 0 | 79.67 | 81.26 | 80.32 |
1 | 78.96 | 80.44 | 80.26 | |
SVM | 0 | 27.58 | 31.23 | 30.47 |
1 | 37.65 | 29.64 | 31.05 | |
KNN | 0 | 42.02 | 43.25 | 42.72 |
1 | 42.02 | 42.07 | 41.78 | |
MLP | 0 | 45.43 | 33.88 | 41.31 |
1 | 42.96 | 36.12 | 39.54 | |
RNN | 0 | 82.91 | 78.26 | 81.85 |
1 | 81.91 | 80.02 | 81.38 | |
AB | 0 | 23.75 | 24.04 | 22.27 |
1 | 24.75 | 26.69 | 27.28 | |
GBDT | 0 | 42.31 | 41.37 | 43.84 |
1 | 41.37 | 40.42 | 43.08 | |
DT | 0 | 78.90 | 77.90 | 78.43 |
1 | 76.72 | 77.67 | 76.90 | |
ET | 0 | 76.72 | 76.72 | 77.14 |
1 | 77.14 | 76.78 | 76.84 | |
VOTE | 0 | 52.80 | 51.74 | 56.28 |
1 | 54.10 | 51.97 | 57.22 | |