Assessing the applicability of game theory and generative adversarial networks in forensics threat detection

The implementation of forensic techniques for password detection has garnered substantial scientific attention recently. Prior studies have explored the detection of forensic attacks on passwords but did not optimise interactions between attackers and defenders. They also failed to accurately detect fake passwords. Addressing these issues, this approach uses appropriate datasets and a novel generative adversarial network (GAN) technique for detecting digital forensic attacks. Integrating game theory and GANs for forensic threat detection enhances robustness and adaptability, enabling proactive defence plans and dynamic threat modelling. This fusion improves the interaction between attackers and defenders and increases the accuracy of false password detection. Utilising the RockYou dataset, the research trains a GAN model to detect forensic attacks. The generator produces new training instances, while the discriminator classifies them. Game theory significantly optimises the generated samples through accurate decision-making, enhancing interaction comfort between attackers and defenders. The proposed framework achieves a prediction accuracy of 97.89%, surpassing existing methods. Consistently enhancing GAN structures could further improve the creation of realistic password patterns, benefiting applications like system security and password authentication.