Table 4 Comparison of the performance in RUTI prediction models in the clinical visit through fivefold cross validation (sample size = 963). Abbreviation: RUTI, recurrent urinary tract infection.

From: Machine learning to predict the development of recurrent urinary tract infection related to single uropathogen, Escherichia coli

Algorithm

Accuracy

Sensitivity

Specificity

Mean

Standard deviation

Mean

Standard deviation

Mean

Standard deviation

Logistic regression

0.604

0.044

0.648

0.117

0.597

0.044

Decision tree

0.654

0.020

0.618

0.058

0.660

0.023

Random forest

0.700

0.039

0.626

0.131

0.712

0.046

Back to article page