Algorithms for Sparse Optimization

Michael Friedlander
University of British Columbia
Computer Science

Many applications in signal processing and machine learning need to solve optimization problems whose solutions are in some sense sparse. The aim is to find the "best" compact representation of some object. Convex optimization plays a key role. The race is on to develop fast algorithms for problems that often are deceptively simple, yet challenge our very best approaches. We will survey some of the main research threads in sparse optimization.

Presentation (PDF File)

Back to Optimization Tutorials