I direct the Artificial Intelligence and Systems Laboratory (AISys). AISys is located at the Storey Innovation Center (Office: 2212).
At AISys, we investigate a variety of open problems that sit at the intersection of artificial intelligence, machine learning, and computer systems (embedded, cloud, robotics). We investigate the development of novel algorithmic and theoretically principled ML methods for systems problems such as optimizing the performance and energy efficiency of highly-configurable systems. We also look into the design and architecture of system software that treat ML computation as a first-class citizen such as optimizing training and inference. Our overarching goal is to develop the next generation of on-device and cloud-based systems able to perceive, reason and react to complex real-world environments and users with high levels of precision and efficiency. The algorithms that we develop use and extend theory from deep learning and neural networks, transfer learning, representation learning, nonconvex optimization, causal learning, reinforcement learning, and robust optimization. We aim to conduct cutting-edge and high impact research through full-stack approaches that encourage lab members with skills in algorithms, systems, statistics, mathematics and software to work closely together to solve critical and practical challenges in the areas at the intersection of AI+Systems.
At AISys, fruitful collborations and constant learning matters a lot to all of us. We have a culture where students frequently collaborate with each other. We typically combine theoretical and empirical insights to build a principled and thorough understanding of key techniques in machine learning, such as deep learning, as well as the challenges we face in this context. Currently, a major theme in our lab is to develop secure, robust, reliable, and performant machine learning systems.
-
Are you a current Ph.D. student at UofSC interested in working with me? The door of my office is always open and I love discussing research with students, please stop by and let’s chat about your interests.
-
Are you a prospective Ph.D. student interested in working with me? I am always looking for highly-motivated students to join our AISys Lab at UofSC. Please drop me an email and share your interests (reading some of our recent publications would help a lot to get an idea what kind of research we have been doing). Apply to the CS PhD program at UofSC.
-
Are you a UofSC undergraduate student interested in working with me on research? Please email me your resume and a description on why you want to get involved. You have a very good chance if: you have contributed to a project in the past (e.g., via a prior experience, or even some personal projects); You have a strong background in computer science and programming (and ideally good in math); You are able to commit 15+ hours/week.
Current Members
Jianhai is a PhD student in computer science. He is interested in transfer learning, autonomic computing, and reinforcement learning.
Shahriar is a PhD student in computer science. He is interested in multi-objective optimization of deep neural networks in embedded systems.
Peter is a PhD student in computer science. He is interested in autonomic computing and reinforcement learning.
Mohammad Ali is a postdoctorate student in computer science. He is interested in probabilistic graphical models, Bayesian networks, and transfer learning.
Blake is a honors college undergraduate student in computer science and math. He is interested in deep neural network model compression and distillation.
Stephen is a honors college undergraduate student in computer science and math. He is interested in deep neural network model compression and distillation.