Many real-life optimization problems need to make decisions with discrete variables and multiple, conflicting objectives. Due to this need, the ability to solve such problems is an important and active area of research. We present a new algorithm, called Pareto Leap, for identifying the (weak) Pareto slices of biobjective mixed-integer programs (BOMIPs), even when Pareto … Read more
The Lovasz theta function is a semidefinite programming (SDP) relaxation for the maximum weighted stable set problem, which is tight for perfect graphs. However, even for perfect graphs, there is no known rounding method guaranteed to extract an optimal stable set from the SDP solution. In this paper, we develop a novel rounding scheme for … Read more
Multiobjective blackbox optimization deals with problems where the objective and constraint functions are the outputs of a numerical simulation. In this context, no derivatives are available, nor can they be approximated by finite differences, which precludes the use of classical gradient-based techniques. The DMulti-MADS algorithm implements a state-of-the-art direct search procedure for multiobjective blackbox optimization … Read more
Inspired by collaborative initiatives in the military domain, we analyze a setting in which multiple different players (e.g., Ministries of Defence) have to carry out preventive maintenance jobs. Each player is responsible for one job, with a job-specific minimal frequency and with maintenance costs, consisting of a job-specific variable component and a fixed component, which … Read more
Nonlinear matrix decomposition (NMD) with the ReLU function, denoted ReLU-NMD, is the following problem: given a sparse, nonnegative matrix \(X\) and a factorization rank \(r\), identify a rank-\(r\) matrix \(\Theta\) such that \(X\approx \max(0,\Theta)\). This decomposition finds application in data compression, matrix completion with entries missing not at random, and manifold learning. The standard ReLU-NMD … Read more
Being inspired by the parametric decomposition theorem for multiobjective optimization problems (MOPs) of Cuenca Mira and Miguel García (2017), and by the block-coordinate descent for single objective optimization problems, we present a decomposition theorem for computing the set of minimal elements of a partially ordered set. This set is decomposed into subsets whose minimal elements … Read more
In this paper, we address a challenging problem faced by a Brazilian oil and gas company regarding the rescheduling of helicopter flights from an onshore airport to maritime units, crucial for transporting company employees. The problem arises due to unforeseen events like bad weather or mechanical failures, leading to delays or postponements in the original … Read more
We investigate a multi-stage version of the selection problem where the variation between solutions in consecutive stages is either penalized in the objective function or bounded by hard constraints. While the former problem turns out to be tractable, the complexity of the latter problem depends on the type of bounds imposed: When bounding the number … Read more
The representation of a function in a higher-dimensional space is often referred to as lifting. Liftings can be used to reduce complexity. We are interested in the question of how liftings affect the local convergence of Newton’s method. We propose algorithms to construct liftings that potentially reduce the number of iterations via analysis of local … Read more
An algorithm based on the interior-point methodology for solving continuous nonlinearly constrained optimization problems is proposed, analyzed, and tested. The distinguishing feature of the algorithm is that it presumes that only noisy values of the objective and constraint functions and their first-order derivatives are available. The algorithm is based on a combination of a previously … Read more