Table 10 Comparison of accuracy for different machine learning methods across the 16 cases.
From: Evaluation of liquefaction potential in central Taiwan using random forest method
Case | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |
|---|---|---|---|---|---|---|---|---|---|
Accuracy (%) | DNN | 96.85 | 96.67 | 95.19 | 96.30 | 90.37 | 95.00 | 93.15 | 93.70 |
RF | 98.89 | 98.34 | 97.60 | 98.71 | 96.67 | 98.52 | 96.86 | 97.63 | |
SVM | 95.74 | 95.74 | 92.59 | 96.11 | 90.37 | 95.74 | 90.37 | 95.19 | |
LSTM | 99.07 | 98.52 | 98.52 | 73.52 | 92.96 | 97.41 | 95.74 | 70.89 | |
Case | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | |
|---|---|---|---|---|---|---|---|---|---|
Accuracy (%) | DNN | 90.19 | 94.44 | 93.89 | 89.63 | 94.63 | 89.81 | 94.07 | 89.63 |
RF | 95.75 | 98.34 | 96.67 | 95.19 | 97.60 | 95.75 | 96.67 | 95.19 | |
SVM | 82.96 | 95.19 | 92.59 | 87.04 | 90.37 | 88.15 | 95.00 | 90.37 | |
LSTM | 90.37 | 97.41 | 72.56 | 90.19 | 91.30 | 73.52 | 70.89 | 89.63 | |
