MaTais Caldwell

MaTais (Taz) Caldwell is a PhD candidate in Computer Science at the University of Mississippi, where his research focuses on generative adversarial networks (GANs) and deep learning. His current work centers on EmbeddGAN, a novel GAN framework that replaces the traditional discriminator with an embedding network trained using Gini distance correlation, offering a principled and stable alternative to conventional adversarial training. He is also developing cEmbeddGAN, a conditional extension of EmbeddGAN with an emphasis on label-conditioned image generation. His research sits at the intersection of machine learning theory, representation learning, and dependable AI systems. 

 

Taz has taught an introductory course in computer networking, covering foundational topics such as Application layer protocols (HTTP, SMTP,  DNS, BitTorrent, etc.), and foundations of reliable data transport. He is committed to accessible, hands-on instruction and enjoys mentoring undergraduate students. 

 

Outside of research, Taz loves gaming and enjoys chess. He is an enthusiast of Linux and open-source software, and spends time exploring the intersection of free software culture and everyday computing