Using Autoencoder-Driven Machine Learning for Advance Cybersecurity Malware Detection
Radhey Shyam, G. Chiranjeevi, Rahul Reddy, Ali A. Khala, Swati Sah, Sandeep Kautish, Rejwan Bin Sulaiman · FMDB Transactions on Sustainable Intelligent Networks · 2024
In the present study, we tackle the problem of improving cybersecurity by creating a sophisticated artificial intelligence model that can precisely detect and categorize malware. In this work, we apply an autoencoder algorithm specifically designed to handle the high-dimensional and intricate data the Ember dataset contains. The objective is to develop an artificial intelligence system that can accurately identify benign and malicious executables and advance our knowledge of malware traits. Our goal is to capture the complex representations of the data by using an autoencoder. This will enable a more robust feature learning process, which is necessary to identify advanced cyber threats precisely. Although a thorough explanation of the model’s statistical metrics is saved for the paper’s main body, the abstract refers to encouraging findings that point to the autoencoder’s ability to produce a highly accurate AI for malware classification. This lays the groundwork for a safer online environment by utilizing machine learning to combat new and emerging cyber threats.