Table 1 Accuracy results in a 300 observations proportional hazards, zero skew, 10 and 30% censoring. Prediction accuracy of all tested approaches when simulated data was generated with 300 observations and the following assumptions: proportional hazards, zero skew, 10 and 30% censoring. The table summarizes the mean (and standard deviation) of the following metrics: accuracy, Matthews’ correlation, normalized mutual information (NMI), area under the Receiver Operating Characteristic curve (AUC-ROC), sensitivity (Sn), specificity (Sp) and F1-score (F1)
From: Enhancing SVM for survival data using local invariances and weighting
| Â | 10% censoring | 30% censoring | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Method | Accuracy | Matthews | NMI | AUC-ROC | Sn | Sp | F1 | Accuracy | Matthews | NMI | AUC-ROC | Sn | Sp | F1 |
Cox model | 0.89 (0.02) | 0.78 (0.03) | 0.50 (0.05) | 0.96 (0.01) | 0.61 (0.04) | 0.60 (0.04) | 0.60 (0.04) | 0.89 (0.02) | 0.79 (0.04) | 0.51 (0.06) | 0.96 (0.01) | 0.61 (0.03) | 0.60 (0.04) | 0.60 (0.04) |
Kernel Cox | 0.81 (0.02) | 0.62 (0.05) | 0.30 (0.05) | 0.88 (0.02) | 0.42 (0.01) | 0.91 (0.03) | 0.50 (0.01) | 0.80 (0.02) | 0.59 (0.04) | 0.26 (0.05) | 0.86 (0.02) | 0.32 (0.01) | 0.93 (0.03) | 0.52 (0.01) |
wSVM-KM | 0.75 (0.03) | 0.50 (0.06) | 0.19 (0.05) | 0.87 (0.02) | 0.40 (0.01) | 0.95 (0.03) | 0.54 (0.01) | 0.68 (0.02) | 0.39 (0.05) | 0.11 (0.03) | 0.86 (0.03) | 0.39 (10.25) | 0.95 (0.03) | 0.55 (0.1) |
wSVM-Prop | 0.75 (0.03) | 0.50 (0.06) | 0.18 (0.05) | 0.87 (0.02) | 0.39 (0.01) | 0.95 (0.03) | 0.53 (0.01) | 0.68 (0.02) | 0.38 (0.05) | 0.11 (0.03) | 0.85 (0.02) | 0.39 (0.01) | 0.92 (0.03) | 0.54 (0.1) |
pSVM-linear-KM | 0.88 (0.02) | 0.73 (0.04) | 0.46 (0.05) | 0.95 (0.01) | 0.85 (0.03) | 0.84 (0.03) | 0.81 (0.02) | 0.88 (0.02) | 0.72 (0.05) | 0.43 (0.07) | 0.95 (0.02) | 0.87 (0.04) | 0.87 (0.03) | 0.82 (0.02) |
pSVM-linear-prop | 0.87 (0.02) | 0.73 (0.04) | 0.45 (0.05) | 0.95 (0.01) | 0.85 (0.04) | 0.85 (0.03) | 0.80 (0.03) | 0.86 (0.02) | 0.72 (0.05) | 0.42 (0.07) | 0.94 (0.02) | 0.86 (0.04) | 0.86 (0.03) | 0.80 (0.03) |
pSVM-radial-KM | 0.79 (0.02) | 0.57 (0.05) | 0.25 (0.05) | 0.88 (0.02) | 0.69 (0.07) | 0.88 (0.04) | 0.74 (0.05) | 0.79 (0.02) | 0.58 (0.04) | 0.27 (0.04) | 0.86 (0.02) | 0.67 (0.07) | 0.88 (0.04) | 0.74 (0.05) |
pSVM-radial-prop | 0.77 (0.02) | 0.57 (0.05) | 0.24 (0.05) | 0.88 (0.02) | 0.67 (0.07) | 0.85 (0.04) | 0.72 (0.05) | 0.77 (0.02) | 0.58 (0.04) | 0.27 (0.04) | 0.86 (0.02) | 0.69 (0.03) | 0.87 (0.05) | 0.74 (0.05) |
LUPI-linear-KM | 0.78 (0.03) | 0.56 (0.05) | 0.28 (0.05) | 0.84 (0.03) | 0.81 (0.04) | 0.74 (0.04) | 0.75 (0.03) | 0.77 (0.02) | 0.55 (0.05) | 0.27 (0.06) | 0.84 (0.03) | 0.81 (0.04) | 0.74 (0.04) | 0.77 (0.03) |
LUPI-linear-prop | 0.77 (0.03) | 0.55 (0.05) | 0.28 (0.05) | 0.84 (0.03) | 0.81 (0.04) | 0.74 (0.04) | 0.75 (0.03) | 0.77 (0.02) | 0.55 (0.05) | 0.27 (0.06) | 0.84 (0.03) | 0.81 (0.04) | 0.73 (0.04) | 0.75 (0.03) |
inSVM-gradient | 0.84 (0.02) | 0.68 (0.05) | 0.37 (0.06) | 0.92 (0.02) | 0.87 (0.04) | 0.90 (0.03) | 0.84 (0.04) | 0.80 (0.02) | 0.60 (0.05) | 0.28 (0.05) | 0.89 (0.02) | 0.87 (0.04) | 0.90 (0.03) | 0.84 (0.04) |
inSVM-averaging | 0.83 (0.02) | 0.66 (0.05) | 0.35 (0.06) | 0.92 (0.02) | 0.88 (0.04) | 0.89 (0.03) | 0.84 (0.04) | 0.83 (0.02) | 0.66 (0.05) | 0.35 (0.06) | 0.92 (0.02) | 0.88 (0.04) | 0.89 (0.03) | 0.84 (0.04) |