Robert M. Freund
Robert M. Freund
Publications/Talks/CollaboratorsNew Online Course: Data Science and Analytics
Three Recent Papers
"Using Taylor-approximated Gradients to Improve the Frank-Wolfe Method for Empirical Risk Minimization", with Zikai Xiong, submitted. For additional results and extensions, see [link].
"Analysis of the Frank-Wolfe Method for Convex Composite Optimization involving a Logarithmically-Homogeneous Barrier", with Renbo Zhao, to appear in Mathematical Programming A.
"An 'Oblivious' Ellipsoid Algorithm for Solving a System of (In)Feasible Linear Inequalities," with Jourdain Lamperski and Michael Todd, to appear in Mathematics of Operations Research.
Research Interests
- Nonlinear optimization theory, applications, and computation -- current focus on algorithmic theory and practice of first-order methods
- Computational complexity of nonlinear optimization
- Interior-point methods in convex optimization
- Computational science
- Related mathematical systems
- Applied optimization in management and engineering
- Linear optimization
- Fixed-point methods