<![CDATA[The increasingly complex and dynamic threat of malware drives the need for a more adaptive detection strategy than conventional signature-based methods. This study aims to evaluate the effectiveness of machine learning, heuristics, and big data approaches in detecting modern malware. The main problem raised is the limitation of traditional methods in identifying new malware variants, especially those that use obfuscation techniques such as polymorphism and metamorphism. Using a systematic literature study approach to the 2016-2024 literature from various reputable sources, this study compares the performance of each approach based on accuracy, efficiency, and resistance to adversarial attacks. The results of the analysis show that deep learning models such as the Convolutional Neural Network (CNN) have the highest detection accuracy, while heuristic methods excel in initial detection efficiency, and big data provides advantages in the scalability of real-time detection systems. This study concludes that the hybrid integration of these three approaches has the potential to create a malware detection system that is more adaptive and resilient to cyberattacks, although further empirical validation is still needed for real-world implementation.]]>
Copyrights © 2025
