Cybersecurity

Table 16 Accuracy result (%)

From: Exploring best-matched embedding model and classifier for charging-pile fault diagnosis

Classifiers	Improve average	GloVe	Word2vec	BERT
RF	8.36	87.21	91.81	87.33
SVM	− 7.96	22.72	21.10	21.57
KNN	5.66	47.33	47.42	50.22
MLP	3.17	45.12	42.35	42.65
RNN	10.40	90.95	92.10	91.16
AB	3.50	26.88	25.76	27.94
GBDT	0.74	43.30	42.59	43.86
DT	9.58	85.97	91.87	86.15
ET	10.37	84.85	91.63	85.21
VOTE	4.98	56.62	58.68	60.45

Back to article page