Table 9 Precision and recall metrics.

From: Novel dual gland GAN architecture improves human protein localization classification using salivary and pituitary gland inspired loss functions

Architecture

Precision (↑)

Recall (↑)

F1-Score (↑)

p-value (F1)

Dual-Gland GAN (Ours)

0.872

0.835

0.853

–

Traditional GAN

0.631

0.458

0.531

 < 0.001

DCGAN

0.712

0.582

0.641

 < 0.001

WGAN

0.765

0.673

0.716

 < 0.001

Pix2Pix GAN

0.823

0.764

0.792

0.002

Conditional GAN

0.791

0.712

0.749

 < 0.001

CycleGAN

0.812

0.752

0.781

0.004

GSIP-GAN

0.841

0.798

0.819

0.018

ECP-IGANN

0.857

0.813

0.834

0.043

MCI-GAN

0.831

0.784

0.807

0.023

Back to article page