Research Article
Bitcoin Theft Detection Based on Supervised Machine Learning Algorithms
Table 3
Effectiveness of test set classification before and after training set oversampling.
| | | Recall (%) | Precision (%) | F1 (%) | | B | C | RC | B | C | RC | B | C | RC |
| | KNN | 84.8 | 92.4 | 8.9 | 87.4 | 72.2 | –17.3 | 86.1 | 81.1 | –5.8 | | SVM | 0.0 | 90.1 | — | 0.0 | 9.0 | — | 0.0 | 16.3 | — | | RF | 92.4 | 95.9 | 3.7 | 98.1 | 95.9 | –2.2 | 95.2 | 95.9 | 0.7 | | AdaBoost | 93.6 | 97.0 | 3.6 | 94.7 | 82.6 | –12.7 | 94.1 | 89.3 | –5.1 | | MLP | 0.0 | 93.0 | — | 0.0 | 57.9 | — | 0.0 | 71.4 | — |
|
|
