Opening Convex Optimization II: L1 Methods for Convex-Cardinality Problems
Convex Optimization II: L1 Methods for Convex-Cardinality Problems
- Length
- 1 h 3 min
- Level
- Graduate
- Recorded
- Stanford University2008
Chapter 1 of 9: Before the lecture
Justifies as the convex envelope of cardinality, then carries it to matrix rank via the nuclear norm.
L1 norm minimization serves as a convex relaxation and heuristic for the minimum cardinality problem of finding the sparsest vector inside a convex set. The lecture examines this relaxation through the convex envelope of the cardinality function and extends the approach to asymmetric and weighted L1 heuristics for bounded variables. It then details applications such as regressor selection, sparse signal reconstruction, total variation denoising for signals and two-dimensional images, and time series change detection. Finally, the framework extends to matrices through the nuclear norm trace heuristic to solve low-rank factor model recovery.
L1 norm minimization serves as a convex relaxation and heuristic for the minimum cardinality problem of finding the sparsest vector inside a convex set. The lecture examines this relaxation through the convex envelope of the cardinality function and extends the approach to asymmetric and weighted L1 heuristics for bounded variables. It then details applications such as regressor selection, sparse signal reconstruction, total variation denoising for signals and two-dimensional images, and time series change detection. Finally, the framework extends to matrices through the nuclear norm trace heuristic to solve low-rank factor model recovery.
BV book

This lecture
Lecture 34 of 37 of Convex Optimization I.
Choose a node to see how it connects to this lecture. What comes before is on the leftat the top. What comes after is on the rightat the bottom. The ring is what it teaches. People and places are above. Other lectures and books are below.
Press a chosen node again to open it.
Your comment is kept while you sign up.Sign in