A COLLABORATIVE FRAMEWORK FOR ANDROID MALWARE IDENTIFICATION USING DYNAMIC ANALYSIS

Thayaaleni, Rajandran (2019) A COLLABORATIVE FRAMEWORK FOR ANDROID MALWARE IDENTIFICATION USING DYNAMIC ANALYSIS. [Final Year Project Report] (Unpublished)

	PDF Thayaaleni Rajandran - 24 pgs.pdf Download (609kB)
	PDF (Please get the password by email to repository@unimas.my , or call ext: 082-583914/3973/3933) Thayaaleni Rajandran.pdf Restricted to Registered users only Download (1MB)

Abstract

The project proposed a dynamic analysis technique in Android malware detection. The objectives of the project are to investigate the Android malware using dynamic analysis technique and to enhance the accuracy of malware detection. The scope of this project focuses on Android Malware detection by using dynamic analysis. The methods to implement this project is through data collection, feature extraction, feature selection, and classification process. The machine learning algorithm is used to train and test datasets with the percentage of 70% which is 140 samples from malware and benign applications and 30% of total datasets which is 60 samples from malware and benign applications respectively. The Correlationbased Feature Selection Evaluator (CfsSubset) algorithm is applied in feature selection process in order to improve the classification process. Lastly, the classification result is generated. The proposed project will extract the features of system calls, network packets, CPU usage and battery usage of the application. The proposed project achieves overall accuracy level of 96.67% using Sequential Minimal Optimization classifier.

Actions (For repository members only: login required)

View Item