We discuss the issue of finding a good mathematical programming solver configuration for a particular instance of a given problem, and we propose a two-phase approach to solve it. In the first phase we learn the relationships between the instance, the configuration and the performance of the configured solver on the given instance. A specific difficulty of learning a good solver configuration is that parameter settings may not all be independent; this requires enforcing (hard) constraints, something that many widely used supervised learning methods cannot natively achieve. We tackle this issue in the second phase of our approach, where we use the learnt information to construct and solve an optimization problem having an explicit representation of the dependency/consistency constraints on the configuration parameter settings. We discuss computational results for two different instantiations of this approach on a unit commitment problem arising in the short-term planning of hydro valleys. We use logistic regression as the supervised learning methodology and consider CPLEX as the solver of interest.

We propose a methodology, based on machine learning and optimization, for selecting a solver configuration for a given instance. First, we employ a set of solved instances and configurations in order to learn a performance function of the solver. Secondly, we formulate a mixed-integer nonlinear program where the objective/constraints explicitly encode the learnt information, and which we solve, upon the arrival of an unknown instance, to find the best solver configuration for that instance, based on the performance function. The main novelty of our approach lies in the fact that the configuration set search problem is formulated as a mathematical program, which allows us to a) enforce hard dependence and compatibility constraints on the configurations, and b) solve it efficiently with off-the-shelf optimization tools.

The alternating current optimal power flow problem is a fundamental problem in the management of smart grids. In this paper we consider a variant which includes activation/deactivation of generators at some of the grid sites. We formulate the problem as a mathematical program, prove its NP-hardness w.r.t. activation/deactivation, and derive two perspective reformulations.

Complex and delocalized manufacturing industries require high levels of integration between production and transportation in order to effectively implement lean and agile operations. There are, however, limitations in research and applications simultaneously embodying further sustainability dimensions. This article presents a methodological framework based on optimization and simulation to integrate (i) aggregate optimized plans for production and multimodal transportation with (ii) detailed dynamic distribution plans affected by demand uncertainty. The objective function of the optimization model considers supply, production, transportation and CO2 emission costs as well as collaboration over the multimodal network. Bill-of-materials and capacity constraints are included. A feedback between simulation and optimization is used to plan requirements for materials and components. Computational experiments are based on realistic instances. Results demonstrate that the framework can be effectively used to analyze cost-CO2 emissions trade-offs, effects of demand uncertainty and collaborative distribution strategies on economic and environmental performance of the supply chain.

In recent years, home care (HC) service systems have been developed as alternatives to conventional hospitalization. Many resources are involved in delivering HC service, including different categories of human resources, support staff, and material resources. One of the main issues encountered while planning human HC resources is the patient assignment problem, i.e., deciding which operator(s) will take care of which admitted patient given some sets of constraints (e.g., the continuity of care). This paper addresses the resource assignment problem for HC systems. A set of mathematical programming models to balance the workloads of the operators within specific categories are proposed. The models consider several peculiarities of HC services, such as the continuity of care constraint, operators' skills, and the geographical areas which patients and operators belong to. Given the high variability of patient demands, models are developed under the assumption that patients' demands are either deterministic or stochastic. The analysis of the results obtained from a real case study demonstrates the applicability of the proposed models as well as the benefits that stem from applying them. Moreover, the obtained results show that an acceptable level of continuity of care cannot be obtained without modeling the continuity of care as a hard constraint. The analysis under continuity of care also shows the high value of information and the difficulties of fully balancing workloads with the application of standard techniques.

The aim of the paper is first of all to describe and formulate some mathematical programming problems, which arise in a railroad company, and which are at present unsolved, at least as it regards large-scale real situation. Problems of such a king are crew and manpower planning, optimal time-table determination. For the crew scheduling problem a new algorithm is proposed which is based on the upper bound linear assignment algorithm. The optimal time-table problem is formulated as an optimal vertex-packing on an undirected graph, with additional linear integer constraints; this integer linear program is structured, i. e., it has a block angular matrix of the constraining system. Computational experience has been made on a sample of real problems coming from the Italian railroad company.