An algorithmic characterization of P-matricity II: adjustments, refinements, and validation

The paper “An algorithmic characterization of P-matricity” (SIAM Journal on Matrix Analysis and Applications, 34:3 (2013) 904–916, by the same authors as here) implicitly assumes that the iterates generated by the Newton-min algorithm for solving a linear complementarity problem of dimension n, which reads 0 ⩽ x ⊥ (M x + q) ⩾ 0, are … Read more

Analysis of Limited-Memory BFGS on a Class of Nonsmooth Convex Functions

The limited memory BFGS (L-BFGS) method is widely used for large-scale unconstrained optimization, but its behavior on nonsmooth problems has received little attention. L-BFGS can be used with or without “scaling”; the use of scaling is normally recommended. A simple special case, when just one BFGS update is stored and used at every iteration, is … Read more

An oracle-based projection and rescaling algorithm for linear semi-infinite feasibility problems and its application to SDP and SOCP

We point out that Chubanov’s oracle-based algorithm for linear programming [5] can be applied almost as it is to linear semi-infinite programming (LSIP). In this note, we describe the details and prove the polynomial complexity of the algorithm based on the real computation model proposed by Blum, Shub and Smale (the BSS model) which is … Read more

Structural Properties of Feasible Bookings in the European Entry-Exit Gas Market System

In this work we analyze the structural properties of the set of feasible bookings in the European entry-exit gas market system. We present formal definitions of feasible bookings and then analyze properties that are important if one wants to optimize over them. Thus, we study whether the sets of feasible nominations and bookings are bounded, … Read more

Analysis of Models for the Stochastic Outpatient Procedure Scheduling Problem

In this paper, we present a new stochastic mixed-integer linear programming model for the Stochastic Outpatient Procedure Scheduling Problem (SOPSP). In this problem, we schedule a day’s worth of procedures for a single provider, where each procedure has a known type and associated probability distribution of random duration. Our objective is to minimize the expectation … Read more

Greedy Systems of Linear Inequalities and Lexicographically Optimal Solutions

The present note reveals the role of the concept of greedy system of linear inequalities played in connection with lexicographically optimal solutions on convex polyhedra and discrete convexity. The lexicographically optimal solutions on convex polyhedra represented by a greedy system of linear inequalities can be obtained by a greedy procedure, a special form of which … Read more

Wasserstein Distributionally Robust Kalman Filtering

We study a distributionally robust mean square error estimation problem over a nonconvex Wasserstein ambiguity set containing only normal distributions. We show that the optimal estimator and the least favorable distribution form a Nash equilibrium. Despite the non-convex nature of the ambiguity set, we prove that the estimation problem is equivalent to a tractable convex … Read more

Rapid prototyping of parallel primal heuristics for domain specific MIPs: Application to maritime inventory routing

Parallel Alternating Criteria Search (PACS) relies on the combination of computer parallelism and Large Neighborhood Searches to attempt to deliver high quality solutions to any generic Mixed-Integer Program (MIP) quickly. While general-purpose primal heuristics are widely used due to their universal application, they are usually outperformed by domain-specific heuristics when optimizing a particular problem class. … Read more

Second-order Guarantees of Distributed gradient Algorithms

We consider distributed smooth nonconvex unconstrained optimization over networks, modeled as a connected graph. We examine the behavior of distributed gradient-based algorithms near strict saddle points. Specifically, we establish that (i) the renowned Distributed Gradient Descent (DGD) algorithm likely converges to a neighborhood of a Second-order Stationary (SoS) solution; and (ii) the more recent class … Read more

Hamiltonian Descent Methods

We propose a family of optimization methods that achieve linear convergence using first-order gradient information and constant step sizes on a class of convex functions much larger than the smooth and strongly convex ones. This larger class includes functions whose second derivatives may be singular or unbounded at their minima. Our methods are discretizations of … Read more