الفهرس 
An Overview of the Android operating systemOnly 14 pages are availabe for public view
 |  | from | 121 |  |  |
| | | from | 121 | | |
Abstract
Malware is the umbrella term that denotes attacking any system by malicious software. During the last few years, the popularity of Android smartphones led to the sneak of several malware applications into different Android markets without any difficulty. As a consequence of this, malware applications have been grown exponentially in the Android markets. Unfortunately, most of these markets suffer from an inability to detect malware, which results in increasing the probability of infecting users’ phones with these malware applications. The challenging issues in the current methods of malware detection in Android systems are the most discriminative features in Android applications as well as the most powerful technique that is used in the detection. Among the familiar machine learning approaches, the support vector machine (SVM), Random Forests (RF), Naïve Bayes techniques (NB) are the most recently applied in malware detection. The presented research focuses on developing an efficient computational framework based on Deep Belief Networks for malware detection. The proposed framework merges static analysis, dynamic analysis, and system calls in feature extraction to achieve the highest accuracy. The evaluation compares the most familiar machine learning approaches that were applied in malware detection with the proposed framework. The obtained results demonstrate that Deep Belief Chapter 1 Introduction Network (DBN) is a strong competitive alternative in malware detection especially when we consider dynamic analysis features. We compare the results according to the different extracted features sets to demonstrate the most discriminative features set in malware detection. The DBN’s highest accuracy is achieved with a hybrid feature set considering static analysis features, dynamic analysis features, and system calls. The obtained results declare that DBN can achieve 99.1% accuracy, 98.7% precision, 1 true positive rate, 99.3% F-measure with the present dataset. These results may change with another dataset. The main drawback for the proposed classifier is the time wasting. Over and above that, we develop our complete static analysis jar which is a Java library, it adopts different efficient methods in an attempt to facilitate and speed static analysis by handling all the Android applications in only one step rather than considering only one application at a time. Moreover, the proposed jar has the ability to check the similarity between two versions of the same application downloaded from different markets.