System Malware Detection on Android Application File Packages Using Heuristic Optimizer through Hybrid Approach EDT-ABO Algorithm

In the age of ubiquitous sensors and intellectual devices, various techniques have been employed to discover diverse categories of malware variants. But, particular areas of malware detection remain unexplored appropriate to the emergence of new malware variations. Many malware detection programs rely on signatures, such as utilizing a hash function to generate a hash value for an executable file containing malware. However, traditional malware detection methods often fall short in detecting the majority of new iterations of existing malware. This research introduces an innovative approach rooted to improve the usability of an existing framework. Because the modern meta-heuristic algorithm optimal selection is built on African Buffalo Optimization Algorithm, we must obtain a novel model i.e., Enhanced Decision Tree Algorithm Using African Buffalo Optimization (EDTABO) to increase the overall optimization. This is accomplished by applying the EDTABO optimisation strategy, which is known as EDTABO to increase its accuracy. This approach is capable of overcoming the complexity provided in engineering issues. The system is initially trained using datasets from malware software and android applications, and the datasets are then pre-processed by removing training errors. The built ABDT model then executes the detection process to foretell malware activity. Also, techniques both static and dynamic are used to detect malware activity. Using such proposed model presented model successfully predicts whether the provided application is malicious or benign and was created for malware detection and protection.