Table 1 Confusion matrix of the CNN classifier for the test set.

From: Breast cancer patient characterisation and visualisation using deep learning and fisher information networks

Test set confusion matrix (72% overall accuracy)

Predicted

Background

Calcification Benign

Calcification Malignant

Mass Benign

Mass Malignant

Actual

Background

0.90

0.04

0.02

0.03

0.02

Calcification Benign

0.30

0.57

0.09

0.01

0.03

Calcification Malignant

0.13

0.26

0.46

0.03

0.11

Mass Benign

0.25

0.02

0.01

0.57

0.15

Mass Malignant

0.10

0.01

0.10

0.12

0.66

  1. The rows are normalised (each sums to one) to show the correctly classified proportion in each class, namely: Background, Calcification Benign, Calcification Malignant, Mass Benign and Mass Malignant.
Back to article page