A Conditional Gradient Framework for Composite Convex Minimization with Applications to Semidefinite Programming
We propose a conditional gradient framework for a composite convex minimization template with broad applications. Our approach combines the notions of smoothing and homotopy under the CGM framework, and provably achieves the optimal O(1/sqrt(k)) convergence rate. We demonstrate that the same rate holds if the linear subproblems are solved approximately with additive or multiplicative error. Specific applications of the framework include the non-smooth minimization semidefinite programming, minimization with linear inclusion constraints over a compact domain. We provide numerical evidence to demonstrate the benefits of the new framework.
hcgm-icml2018.pdf
openaccess
1.65 MB
Adobe PDF
a79fbd590ce1a55a65c1d89db9a22b6a
hcgm-infoscience.pdf
openaccess
1.63 MB
Adobe PDF
47b600f8b53e90d1afb0dc83c3814fbc