Abstract:

There is still new Android spyware being developed despite the widespread adoption of Android applications. Since Android devices account for 72.2% of all smartphone purchases, it stands to reason that they would be a primary target for malware activities. Hackers use a wide variety of techniques to compromise mobile devices, including stealing credentials, wiretapping, and displaying harmful advertisements. Android virus identification has been the subject of extensive study, with numerous experts presenting their hypotheses and methods. Because they can generate a classification from a collection of training examples, ML-based methods have proven useful in detecting these assaults, as they do away with the need for specific signature characterization in malware detection. As part of this effort, we looked closely at how different methods of detecting Android adware use machine learning. Recent studies suggest that machine learning is effective for detecting Android adware, making this an attractive and practical option.

Description:

machine learning, K-Mean