Chi Jin is a PhD candidate in Computer Science at UC Berkeley, advised by Michael Jordan. He received BS in Physics from Peking University in 2012. His research primarily focuses on learning problems and optimization algorithms under non-convex setting. His representative works include guarantees for gradient descent / accelerated gradient descent to efficiently escape saddle points, and the optimization landscape of low-rank problems. His is also recently interested in reinforcement learning problems.
- Foundations of Data Science, Fall 2018. Visiting Graduate Student.