CyberFold

Intrusion detection is a complex and ever-evolving challenge. Can Artificial Intelligence (AI) and Machine Learning (ML) offer new methods to address these challenges more effectively? In a world where cyber threats are constantly growing and diversifying, finding innovative solutions is more crucial than ever. 

CyberFold project explores how Manifold Learning (MnL) algorithms can provide a novel approach to processing traffic and network activity data. These methods enable feature extraction and the generation of low-dimensional embeddings, facilitating domain detection in the generated latent spaces. By studying and adapting these algorithms, the project aims to enhance intrusion detection in current systems, offering advanced solutions for specific cybersecurity challenges.

CyberFold addresses currently open applications: (1) the identification of massive and sudden requests with atypical origins; (2) protocol masking or tunneling; (3) the detection of abnormal traffic between devices; (4) security in smaller-scale IoT scenarios, such as homes or vehicles; and (5) the reduction of false positives in Intrusion Detection Systems (IDS) for industrial devices.