Malware Detection in Android Mobile Platforms using Data Mining Algorithms

  • Mariam Ali AlAli

Student thesis: Master's Thesis


Malware has always been a problem in regards to any technological advances in the software world. Thus, it is to be expected that smart phones are facing the same issues. In this research project, a practical and effective anomaly based malware detection framework is proposed with an emphasis on Android mobile computing platform. Various feature engineering, feature selection, and an ensemble of efficient data mining algorithms, including: J48 decision tree, K-nearest neighbor, Logistic Regression, Multilayer Perceptron Neural Network, Naive Bayes, Random Forest and Support Vector Machines, are used. A dataset consisting of both benign and malicious applications were installed on a device to analyze the behavioral patterns. Data mining techniques were applied to classify the data as benign or malware and each algorithm was assessed to detect which is more suitable to identify malicious software. The results suggest that Random Forest provided the best outcomes thus making it an effective technique for malware detection.
Date of AwardDec 2016
Original languageAmerican English
SupervisorU Zeyar Aung (Supervisor)


  • Malware
  • Malware Detection
  • Cyber Security
  • Technological Advancement
  • Smart Phones
  • Data Mining Algorithms
  • Sustainability
  • Android System & Security.

Cite this