Skip to main content

Table 3 Confusion Matrix and Diagnostic Performance Metrics for Multivariate Adaptive Regression Spline Algorithm

From: Machine learning in medicine: a practical introduction to techniques for data pre-processing, hyperparameter tuning, and model comparison

 

Reference

 

Prediction

Malignant

Benign

Malignant

79

28

Benign

9

75

Sensitivity

89.8% (79 of 88)

 

Specificity

72.8% (75 of 103)

 

Negative predicitive value

89.3% (75 of 84)

 

Positive predictive value

73.8% (79 of 107)

 

Accuracy

81.2% (155 of 191)

 

No information rate

0.54

 

p-Value [Acc > NIR]*

< 0.001

 
  1. *Refers to a one-sided binomial test determining whether the accuracy proportion is higher than the no-information rate