Genetic Algorithm-Based Feature Optimization for Enhanced Android Malware Detection Using Machine Learning

The Android platform, benefiting from its open-source nature and support from Google, holds the largest global market share. As the world's most widely used operating system, it has attracted the attention of cyber criminals who are primarily targeting users through the widespread dissemination of malicious software. This project introduces a practical machine-learning methodology for identifying Android malware, utilizing an evolutionary genetic algorithm to select distinctive features effectively. Features from the Genetic algorithm are incorporated into training machine learning classifiers, and a comparison is made of their ability to identify Malware both before and after feature selection is applied. The experimental outcomes confirm that the genetic algorithm yields the most optimized feature subset, resulting in a reduction of the feature dimension to less than half of the original feature set. Machine learning classifiers achieve a classification accuracy of over 94% after feature selection, with a significant reduction in the number of features used, resulting in a positive effect on the computational complexity of the learning process