Table 2 The performance metrics of the proposed model CSHG-CervixNet using hold-out validation.

From: A hybrid compound scaling hypergraph neural network for robust cervical cancer subtype classification using whole slide cytology images

Performance metrics- hold out

Specificity

Precision

Recall

F1-score

Dyskeratotic

97.2

98.6

97.9

99.1

Koilocytotic

97.7

93.5

96

94.7

Metaplastic

99.1

97.3

96.1

96.7

Parabasal

100

100

100

100

Superficial_intermediate

100

93.9

96.9

100