Research/Earlier Research
Optimization
under uncertainty.
Research experience in Professor Ying Cui’s group at UC Berkeley.
- Group
- Professor Ying Cui, Department of Industrial Engineering and Operations Research, UC Berkeley
- Role
- Collaborative research
- Tools
- Julia
Schematic · not dataSolving risk-aware problems that are not convex.
Risk-aware objectives make an optimization problem harder, and nonconvexity removes the usual guarantees. The question was how a second-order method behaves there.
A semismooth Newton method, against other solvers.
I deployed a semismooth Newton method to solve a risk-aware optimization model involving a neural network, and compared its running time and solutions with other solvers, especially in the nonconvex case. The algorithms were implemented in Julia.
What this page does not show.
- No results are shown here.
- The figure is a schematic of iterates on level sets, not results.