Android Malware Detection via Hybrid Lion-Bee Optimization with Bi-Directional Recurrent Neural Network

The number of malicious programs that target the Android OS has significantly expanded with the rise in mobile device usage. It is crucial to identify, prevent, and defeat Android malware assaults since they have long presented a severe threat to Android apps. The emergence of secure Android app ecosystem depends on recognizing and classifying harmful applications into groups are similar to another. Malware families can be categorized in order to recognize harmful activity and to systematically spot dangerous-patterns. Thus, the study suggests a hybrid strategy that combines the attributes discovered from doing static and dynamic malware assessment for enhanced identification and categorization of Android malware. This method more effectively addresses the issue of studying, identifying Android malware. The feature extracted from image sections utilizing the hybrid lion and artificial bee colony (HL-ABC) optimization. RNN were used to classify the retrieved features. Each segment of a malware image file served as a test case for the categorization performance. This is contrasted with other current ways to show the effectiveness of the suggested strategy. Comparing the suggested model to other approaches, the study's findings demonstrate its effectiveness in identifying and classifying Android malware.