Diagnosis of Thyroid Nodules Based on Image Enhancement and Deep Neural Networks

Table 4

Performance comparison of different classification methods.

Method	K-nearest neighbors	Decision tree	Naive Bayesian model	Support vector machine	Logistic regression	Reinforcement learning		Deep neural networks
						Hard voting	Soft voting
Accuracy	0.852941	0.794118	0.411765	0.823529	0.823529	0.823529	0.843137	0.901961
Precision	0.857143	0.847826	0.329412	0.868132	0.860215	0.860215	0.841584	0.894737
Recall	0.988235	0.917647	0.903226	0.929412	0.941176	0.941176	1	1
F1-score	0.918033	0.881356	0.482759	0.897727	0.898876	0.898876	0.913978	0.944444