PHANTOM: Progressive High-Fidelity Adversarial Network for Threat Object Modeling

The scarcity of cyberattack data hinders the development of robust intrusion detection systems. This paper introduces PHANTOM, a novel adversarial variational framework for generating high-fidelity synthetic attack data. Its innovations include progressive training, a dual-path VAE-GAN...