📌 **Project Title:** Android Malware Classification Using Machine Learning
📌 **Dataset Used:** TUANDROMD (Tezpur University Android Malware Dataset)
📌 **Techniques Applied:** Static analysis, Random Forest, Logistic Regression
📌 **Models Compared:**
– Random Forest (Non-linear, 99% accuracy)
– Logistic Regression (Linear baseline, 98% accuracy)
🔍 **Contents Covered in the Presentation:**
• Research question and motivation
• Dataset introduction (4,464 Android applications)
• Feature types extracted from AndroidManifest.xml
• Preprocessing and 60/40 train-test split
• Model descriptions and evaluation metrics
• Confusion matrix, ROC curve, feature importance
• Key findings and what was learned
• Limitations and future improvements
🎯 **Key Takeaway:**
Machine learning, particularly Random Forest, can accurately classify Android malware using static manifest-based features.
source