/Malware-detection-ML

Android malware detection using machine learning.

Primary LanguageJupyter NotebookMIT LicenseMIT

Malware-detection-ML

Android malware detection using Machine Learning.

Dataset - NATICUSdroid (Android Permissions)

The notebook 'main.ipynb' contains the following outlines.

  • Import modules
  • Read Data
  • Pre processing
  • Feature Selection methods
    • Variance Inflation Factor (multi-collinearity removal)
    • Mutual Information Score
  • Machine learning pipeline modelling
    • Logistic Regression
    • Support Vector Machines
    • K Nearest Neighbors
    • Random Forest
    • XGBoost
    • CatBoost
    • Voting classifier : Random forest + XGBoost + Catboost
    • Stacking classifier : Logistic regression + SVM + Knn + Random forest + XGBoost + Catboost
  • Bayesian based hyper-parameter tuning
  • ML model explainability using Shapley values

Dataset citations

  • Mathur, Akshay & Mathur, Akshay. (2022). NATICUSdroid (Android Permissions) Dataset. UCI Machine Learning Repository.