Table 3 Comparison of machine learning model performance for predicting SIB.

From: Entropy-based risk network identification in adolescent self-injurious behavior using machine learning and network analysis

Model

Accuracy

AUC

Sensitivity

PPV

Brier Score

RF

0.748

0.814

0.449

0.652

0.160

HGB

0.735

0.800

0.423

0.624

0.168

SVM

0.738

0.802

0.398

0.645

0.167

MLP

0.745

0.808

0.469

0.634

0.164

LGBM

0.743

0.793

0.449

0.638

0.168

KNN

0.728

0.757

0.378

0.622

0.181

  1. RF random forest, HGB HistGradientBoosting, SVM support vector machine, MLP multi-layer perceptron, LGBM LightGBM, KNN K-nearest neighbors.
  2. Metrics include Accuracy, AUC area under the ROC curve, Sensitivity, PPV positive predictive value, Brier Score lower indicates better accuracy.
Back to article page