Automatic Classification of Specific Melanocytic Lesions Using Artificial Intelligence

Table 2

The assessment of classifiers for a 1-vs-all and overall classification.

Measure	Logit	Tree	3NN	5NN	10NN	15NN	SVM
	0.9730	0.8485	0.7200	0.7660	0.6939	0.6735	1.0000
	1.0000	0.9266	1.0000	1.0000	0.9785	0.9677	1.0000
	1.0000	0.9675	0.9914	0.9889	0.9679	0.9659	1.0000
	0.8611	0.6977	0.8519	0.7586	0.8148	0.8182	0.9143
	0.9528	0.9394	0.8870	0.8761	0.8783	0.8500	0.9626
	0.9817	0.9279	0.9615	0.9089	0.9099	0.9030	0.9851
	0.8387	0.9655	0.7419	0.6786	0.6071	0.5185	0.8611
	0.9189	0.9381	0.8919	0.8596	0.8421	0.8174	0.9623
	0.9469	0.9555	0.9152	0.8746	0.7860	0.7378	0.9688
	0.8421	0.7568	0.8235	0.7105	0.6053	0.5227	0.9429
	0.9712	0.9333	0.9352	0.9231	0.8846	0.8776	0.9813
	0.9648	0.9425	0.9606	0.9395	0.8812	0.8545	0.9789
	0.8803	0.8028	0.7746	0.7324	0.6761	0.6197	0.9296

BN: blue nevus, CN: Clark nevus, MM: malignant melanoma, SN: Spitz nevus, ALL: multiclass classification.