Clarabel: An Interior Point Solver for Quadratic Conic Optimization (new room ME C2 405)

Abstract: Convex optimization finds applications in various fields, such as machine learning, signal processing, control systems, finance, logistics, and operations research, and relies on numerical software to solve problems efficiently. We are going to introduce our numerical solver called Clarabel, which is based on an interior point method with homogeneous embedding. It supports a variety of conic constraints beyond linear programming (LP) and quadratic programming (QP), including second-order cones, semidefinite cones, exponential cones, and power cones. The solver was originally developed in Julia and Rust, but we also provide support for Python. It performs faster and more stably than state-of-the-art solvers on a class of conic optimization problems.

 

Bio: Yuwen Chen holds a B.Sc. degree in Electrical Engineering from Shanghai Jiao Tong University (China, 2017) and an M.Sc. degree from the Department of Information Technology and Electrical Engineering at ETH Zurich (Switzerland, 2020). He is currently a third-year Ph.D. student under the supervision of Prof. Paul Goulart in the Department of Engineering Science at the University of Oxford. His research focuses on optimization algorithms and software development for convex optimization.