Self-Concordant Analysis of Frank-Wolfe Algorithms

Pavel Dvurechenskii · Petr Ostroukhov · Kamil Safin · Shimrit Shtern · Mathias Staudigl


Keywords: [ Convex Optimization ] [ Large Scale Learning and Big Data ] [ Optimization - Convex ]

[ Abstract ]
[ Slides
Tue 14 Jul 7 a.m. PDT — 7:45 a.m. PDT
Tue 14 Jul 8 p.m. PDT — 8:45 p.m. PDT


Projection-free optimization via different variants of the Frank-Wolfe (FW), a.k.a. Conditional Gradient method has become one of the cornerstones in optimization for machine learning since in many cases the linear minimization oracle is much cheaper to implement than projections and some sparsity needs to be preserved. In a number of applications, e.g. Poisson inverse problems or quantum state tomography, the loss is given by a self-concordant (SC) function having unbounded curvature, implying absence of theoretical guarantees for the existing FW methods. We use the theory of SC functions to provide a new adaptive step size for FW methods and prove global convergence rate O(1/k) after k iterations. If the problem admits a stronger local linear minimization oracle, we construct a novel FW method with linear convergence rate for SC functions.

Chat is not available.