Table 9 CoLog in comparison to other log anomaly detection methods on the Hadoop dataset. Bold numbers indicate the outstanding results.
Anomaly detection technique | Precision | Recall | F1-Score | Accuracy |
|---|---|---|---|---|
Supervised Methods | ||||
Logistic Regression143 | 48.523 | 50.000 | 49.250 | 97.046 |
Support Vector Machines109 | 48.523 | 50.000 | 49.250 | 97.046 |
Decision Tree110 | 48.523 | 50.000 | 49.250 | 97.046 |
Attentional BiLSTM63 | 97.640 | 97.955 | 97.792 | 97.902 |
Convolutional Neural Network46 | 99.719 | 99.847 | 99.783 | 99.955 |
pylogsentiment38 | 99.886 | 99.732 | 99.809 | 99.905 |
Unsupervised Methods | ||||
Isolation Forest144 | 47.702 | 50.000 | 48.824 | 54.034 |
Principal Component Analysis12 | 49.995 | 49.996 | 49.996 | 58.214 |
LSTM36 | 99.850 | 97.397 | 98.589 | 99.715 |
Transformer80 | 97.280 | 99.833 | 98.518 | 99.685 |
CoLog | \(\mathbf {99.997}\) | \(\mathbf {99.956}\) | \(\mathbf {99.977}\) | \(\mathbf {99.994}\) |
