Skip to main content

Table 4 Relative performance of machine learning algorithm across all therapy-recommendation combinations

From: Computational prediction of multidisciplinary team decision-making for adjuvant breast cancer drug therapies: a machine learning approach

Algorithm

Median rank

Multiclass ADTree (Bagged)

2.0

J48 decision tree (Bagged)

3.0

Ripple down rules (Bagged)

3.0

Multiclass ADTree

5.0

SVM, polynomial kernel (Bagged)

6.0

SVM, radial basis function kernel (Bagged)

7.0

SVM, polynomial kernel

7.0

SVM, radial basis function kernel

9.0

Naive Bayes classifier (Bagged)

9.0

Logistic regression (Bagged)

10.0

Naive Bayes classifier

11.0

J48 decision tree

11.0

Decision table

12.0

Logistic regression

14.0

Nearest neighbour classifier

14.0

Ripple down rules

16.0

OneR (Bagged)

17.0

OneR

17.0

  1. Abbreviations: Bagged bootstrap-aggregated, ADTree alternating decision tree, SVM support vector machine