Table 3 Ten-fold cross-validation results when using Strand with 64-bit hash codes and Asm sequences to predict 10 folds from the Microsoft malware training data
From: Polymorphic malware detection using sequence classification methods and ensembles
10-fold cross-validation results ASM 64-Bit | ||||||
|---|---|---|---|---|---|---|
Fold | Classified | Correct | Sensitivity | Precision | Training | Prediction |
1 | 1087 | 1071 | 98.53% | 99.17% | 0:40:21 | 0:04:19 |
2 | 1087 | 1071 | 98.53% | 98.98% | 0:39:11 | 0:05:31 |
3 | 1087 | 1073 | 98.71% | 99.17% | 0:39:45 | 0:05:19 |
4 | 1087 | 1061 | 97.61% | 98.61% | 0:40:44 | 0:05:24 |
5 | 1087 | 1072 | 98.62% | 99.17% | 0:37:12 | 0:04:49 |
6 | 1087 | 1074 | 98.80% | 99.26% | 0:40:21 | 0:05:12 |
7 | 1087 | 1076 | 98.99% | 99.35% | 0:41:51 | 0:04:33 |
8 | 1087 | 1076 | 98.99% | 99.26% | 0:31:06 | 0:04:36 |
9 | 1086 | 1071 | 98.62% | 98.89% | 0:38:43 | 0:05:08 |
10 | 1086 | 1070 | 98.53% | 98.89% | 0:33:34 | 0:04:11 |