Title:

Tail-GAN: Generating Synthetic Cryptocurrency Time-Series Data with AI.

I designed and trained a Tail-GAN model to generate synthetic cryptocurrency log-return time series that mimic real-world patterns like volatility clustering and fat tails. The project compared Tail-GAN with other models (VAE, Attention-based GANs) using statistical metrics and tail-risk indicators such as Value at Risk (VaR) and Expected Shortfall (ES). This work helps improve financial risk modeling and trading strategy evaluation through realistic data simulation.

This project reflects my passion for building AI systems that model complex, real-world behavior — combining research precision with practical insight.

🔗 GitHub Repository: https://github.com/farid-doy10/TailGAN-Crypto-Dissertation/blob/main/README.md