Research Article
AndroClass: An Effective Method to Classify Android Applications by Applying Deep Neural Networks to Comprehensive Features
Table 1
A summary on specifications of AndroClass and existing methods.
| | Snaz. | LACTA | Ghorb. | Zhu. | Wang. | ClassifyDroid | AndroClass |
| Number of features | 5 | 1 | 1 | 6 | 11 | 1 | 14 | Number of employed datasets | 1 | 1 | 1 | 1 | 1 | 1 | 3 | Utilizing stable features | ✓ | ✓ | ✓ | ✘ | ✓ | ✓ | ✓ | Features representing actual fn. of apps | ✘ | ✓ | ✓ | ✘ | ✓ | ✓ | ✓ | Posing issues to user privacy | ✘ | ✘ | ✘ | ✓ | ✘ | ✘ | ✘ | Classifying un/newly released apps | ✘ | ✓ | ✓ | ✘ | ✓ | ✓ | ✓ |
|
|