Table 8 Comparative analysis of precision, recall and F-measure for existing CCF + BPF and proposed NSVM-CNN

From: An effective image retrieval based on optimized genetic algorithm utilized a novel SVM-based convolutional neural network classifier

No of image retrieved	Flickr Logos 27 dataset
	Precision		Recall		F-measure
	CCF + BPF	NSVM-CNN	CCF + BPF	NSVM-CNN	CCF + BPF	NSVM-CNN
1–4, 6–7, 10	98.87	99.78	58	70	73.11098	82.27824
20	96.11	98.54	62	74.25	75.37562	84.68771
30	93.73	97.97	64.5	77.98	76.41516	86.83945
40	91.09	96.73	68	79.25	77.86938	87.12186
50	88.52	95.82	71.25	81.24	78.95162	87.92971
60	85.95	94.85	74.5	85.88	79.81645	90.1424
70	83.38	93.81	77.75	87.53	80.46664	90.56126
80	82.81	92.99	81	92.18	81.895	92.58323
90	81.24	91.37	84.25	96.83	82.71763	94.0208
100	80	85.78	82	99.69	80.98765	92.21338

Back to article page