Cervical Cancer Diagnosis Model Using Extreme Gradient Boosting and Bioinspired Firefly Optimization

Table 11

Comparison of the proposed study with benchmark studies.

Target class	Model	Accuracy	Sensitivity	Specificity	PPA	NPA
Hinselmann (15)	Abdoh et al. [5]	93.97	100	89.96	84.97	100
	Wu and Zhou [4]	97.6	96.65	98.54	98.48	96.78
	Proposed study	98.38	97.55	99.2	99.17	97.63
Schiller (13)	Abdoh et al. [5]	90.18	98.73	84.63	80.75	99.03
	Wu and Zhou [4]	95.01	93.24	97.58 (12)	97.29 (12)	93.81
	Proposed study	96.39	95.9	96.92	97.1	95.65
Cytology (11)	Abdoh et al. [5]	92.75	100	87.92	83	100
	Wu and Zhou [4]	96.94	95.58 (8)	99.01	98.94	95.76 (8)
	Proposed study	96.98	95.83	99.74	98.73	95.93
Biopsy (11)	Abdoh et al. [5]	94.13	100	90.21	86.07	100
	Wu and Zhou [4]	96.06	94.94 (6)	97.76 (11)	97.58 (11)	94.91
	Proposed study	96.93	95.26	98.73	98.77	95.1