Table 5 Metrics average and standard deviation over 5 random test splits for the performance evaluation on LGG dataset
From: Supervised multiple kernel learning approaches for multi-omics data integration
LGG | |||
|---|---|---|---|
Algorithm | ACC | AUC | F1 |
block PLSDA | \(0.651 \pm 0.024\) | \(0.713 \pm 0.034\) | \(0.677 \pm 0.029\) |
block sPLSDA | \(0.637 \pm 0.030\) | \(0.771 \pm 0.039\) | \(0.692 \pm 0.027\) |
SVM concat | \(\mathbf {0.723 \pm 0.030}\) | \(\mathbf {0.781 \pm 0.024}\) | \(\mathbf {0.741 \pm 0.032}\) |
SVM naive | \(0.709 \pm 0.011\) | \(0.774 \pm 0.024\) | \(0.724 \pm 0.022\) |
SimpleMKL-SVM | \(0.684 \pm 0.011\) | \(0.759 \pm 0.024\) | \(0.710 \pm 0.020\) |
SEMKL-SVM | \(0.691 \pm 0.011\) | \(0.762 \pm 0.028\) | \(0.719 \pm 0.017\) |
STATIS-UMKL + SVM | \(0.709 \pm 0.009\) | \(0.774 \pm 0.023\) | \(0.728 \pm 0.015\) |
Deep MKL (weighted sum) | \(0.687\pm 0.011\) | \(0.765 \pm 0.025\) | \(0.684 \pm 0.031\) |
Cross-Modal Deep MKL (weighted sum) | \(0.700 \pm 0.020\) | \(0.768 \pm 0.026\) | \(0.695 \pm 0.032\) |
NN_VCDN | \(0.703 \pm 0.036\) | \(0.754 \pm 0.030\) | \(0.715 \pm 0.028\) |
Dynamics | \(0.707 \pm 0.029\) | \(0.769 \pm 0.027\) | \(0.714 \pm 0.023\) |
MOGONET | \(0.669 \pm 0.026\) | \(0.711 \pm 0.026\) | \(0.69 \pm 0.032\) |