Table 6 Model performance of the k-nearest neighbors.

From: Identification of perception gaps between physicians and patients with neurological diseases and the prediction of these gaps using machine learning

 

PSQ-18

SDM-Q-9/SDM-Q-Doc

Barthel Index

Original questionnaire

SF-36 subdomains

Holdout log loss

0.1342

0.1709

0.0581

0.1333

0.0500

MCC

0.8100

0.1600

0.2500

0.4200

0.7600

AUC–ROC

0.9888

0.9846

0.9954

0.9907

1.0000

Accuracy

0.9697

0.9444

0.9697

0.9444

1.0000

Precision

1.0000

0.9000

0.9000

0.8947

1.0000

Sensitivity

0.9310

1.0000

1.0000

1.0000

1.0000

Specificity

1.0000

0.8889

0.9583

0.8947

1.0000

  1. AUC, area under the curve; MCC, Matthew’s correlation coefficient; PSQ-18, Patient Satisfaction Questionnaire-18; ROC, receiver operating characteristic; SDM-Q-9, 9-item Shared Decision Making Questionnaire from the patient perspective; SDM-Q-Doc, Shared Decision Making Questionnaire from the physician perspective; SF-36, 36-item Short Form.
Back to article page