We consider the oracle complexity of constrained convex optimization given access to a Linear Minimization Oracle (LMO) for the constraint set and a gradient oracle for the $L$-smooth, $L$-strongly convex objective. This model includes Frank-Wolfe methods and their many variants. Over the problem class of $\alpha$-strongly convex constraint sets $S$, we demonstrate that one can … Read more
This paper studies a class of double-loop (inner-outer) algorithms for convex composite optimization. For unconstrained problems, we develop a restarted accelerated composite gradient method that attains the optimal first-order complexity in both the convex and strongly convex settings. For linearly constrained problems, we introduce inexact augmented Lagrangian methods, including a basic method and an outer-accelerated … Read more
Composite functions have been studied for over 40 years and appear in a wide range of optimization problems. Convex analysis of these functions focuses on (i) conditions for convexity of the function based on properties of its components, (ii) formulas for the convex conjugate of the function based on those of its components and (iii) … Read more
Deploying mathematical optimization in autonomous production systems requires precise contracts for objects returned by an optimization solver. Unfortunately, conventions on dual solution and infeasibility certificates (rays) vary widely across solvers and classes of problems. This paper presents the theoretical framework used by MathOpt (a domain-specific language developed and used at Google) to unify these notions. … Read more
In convex geometry, the Shapley–Folkman Lemma asserts that the nonconvexity of a Minkowski sum of $n$-dimensional bounded nonconvex sets does not accumulate once the number of summands exceeds the dimension $n$, and thus the sum becomes approximately convex. Originally published by Starr in the context of quasi-equilibrium in nonconvex market models in economics, the lemma … Read more
We extend the standard notion of self-concordance to non-convex optimization and develop a family of second-order algorithms with global convergence guarantees. In particular, two function classes – weakly self-concordant functions and F-based self-concordant functions – generalize the self-concordant framework beyond convexity, without assuming the Lipschitz continuity of the gradient or Hessian. For these function classes, … Read more
We propose a simple proof of the worst-case iteration complexity for the Difference of Convex functions Algorithm (DCA) for unconstrained minimization, showing that the global rate of convergence of the norm of the objective function’s gradients at the iterates converge to zero like $o(1/k)$. A small example is also provided indicating that the rate cannot … Read more
This work presents a novel optimization approach for training linear classifiers in multiclass classification tasks, when focusing on a regularized and smooth Weston-Watkins support vector machine (SVM) model. We propose a Majorization-Minimization (MM) algorithm to solve the resulting, Lipschitz-differentiable, optimization problem. To enhance scalability of the algorithm when tackling large datasets, we introduce an incremental … Read more
In this paper, we propose a new acceleration strategy for gradient-based methods applied to strictly convex Quadratic Programming (QP) problems. The strategy consists in performing, at selected iterations, minimization steps along alternative descent directions or even within low-dimensional affine subspaces. In particular, considering the contribution of the linear and quadratic part of the objective function … Read more
Suppose we wish to determine the quality of a candidate solution to a convex stochastic program in which the objective function is a statistical functional parameterized by the decision variable and known deterministic constraints may be present. Inspired by stopping criteria in primal-dual and interior-point methods, we develop cancellation theorems that characterize the convergence of … Read more