Table 3 Data size and division in the existing literature
From: Machine learning methods for predicting residual strength in corroded oil and gas steel pipes
Source | Data size | Ratio of training set | Ratio of validation set | Ratio of test set |
|---|---|---|---|---|
115 | 80% | 10% | 10% | |
254 | 50% | 15% | 35% | |
550 | 80% | - | 20% | |
1815 | 70% | 15% | 15% | |
- | 70% | 15% | 15% | |
453 | 90% | - | 10% | |
150 | 66.4% | 16.6% | 17% | |
292 | 96.6% | - | 3.4% | |
688 | 97.8% | - | 2.2% | |
129 | 84.5% | - | 15.5% | |
25 | 76% | 12% | 12% | |
45 | 80% | 10% | 10% | |
257 | 70% | 15% | 15% | |
75 | 93.3% | - | 6.7% | |
453 | 90% | - | 10% | |
314 | 70% | - | 30% | |
453 | 90% | - | 10% | |
90 | 78.9% | 15.6% | 5.5% | |
217 | 80% | - | 20% | |
39 | 77% | - | 23% | |
572 | 70% | - | 30% | |
61 | 81.9% | - | 18.1% | |
79 | 83.5% | - | 16.5% | |
91 | 89% | - | 11% | |
1353 | 70% | 15% | 15% | |
1843 | 70% | 15% | 15% | |
- | 70% | 15% | 15% | |
100 | 80% | - | 20% | |
193 | 80% | - | 20% | |
453 | 80% | - | 20% |
