Abstract
We study a setting in which a company not only has a fleet of capacitated vehicles and drivers available to make deliveries but may also use the services of occasional drivers (ODs) willing to make deliveries using their own vehicles in return for a small fee. Under such a business model, a.k.a crowdshipping, the company seeks to make all the deliveries at the minimum total cost, i.e., the cost associated with their vehicles plus the compensation paid to the ODs.We consider a stochastic and dynamic last-mile delivery environment in which customer delivery orders, as well as ODs available for deliveries, arrive randomly throughout the day, within fixed time windows.We present a novel deep reinforcement learning (DRL) approach to the problem that can deal with large problem instances. We formulate the action selection problem as a mixed-integer optimization program.The DRL approach is compared against other optimization under uncertainty approaches, namely, sample -average approximation (SAA) and distributionally robust optimization (DRO). The results show the effective-ness of the DRL approach by examining out-of-sample performance.

Abstract
Optimisation problems in Healthcare have received considerable attention for more than three decades. More recently, however, with decreasing birth rates in nearly all of the developed countries and increasing average longevity globally, optimisation issues in Healthcare have become noticeably important and attract keen interest from the Operations Research community. Over the years, attention has gradually expanded from resource allocation and strategic planning to include operational issues such as resource scheduling and treatment planning. This paper surveys several applications of Operations Research in the domain of Healthcare. In particular, the paper reviews key contributions addressing contemporary optimisation issues in this area. It highlights current research activities, focusing on a variety of optimisation problems as well as solution techniques used for solving the optimisation problems.

Abstract
Although single objective metaheuristics are widely spread and applied in many combinatorial optimisation problems, only very recently have multiobjective metaheuristics (MOMH) been designed and used in practice. They aim at obtaining good approximations of the set of nondominated solutions of a problem, in an efficient way. In this work, we have applied multiobjective versions of simulated annealing and taboo search to the resource constrained project scheduling problem (RCPSP), in order to minimise the makespan, the "weighted" lateness of activities and the violation of resource constraints. Computational experience performed on randomly generated instances shows that this general approach is flexible, effective and able to deal with multiple objectives and with variations in the problem structure.

Abstract
In this paper, the Unit Commitment (UC) problem is presented and solved, following an innovative approach based on a metaheuristic procedure. The problem consists on deciding which electric generators must be committed, over a given planning horizon, and on defining the production levels that are required for each generator, so that load and spinning reserve requirements are verified, at minimum production costs. Due to its complexity, exact methods proved to be inefficient when real size problems were considered. Therefore, heuristic methods have for long been developed and, in recent years, metaheuristics have also been applied with some success to the problem. Methods like Simulated Annealing, Tabu Search and Evolutionary Programming can be found in several papers, presenting results that are sufficiently interesting to justify further research in the area. In this paper, a resolution framework based on GRASP - Greedy Randomized Adaptive Search Procedure - is presented. To obtain a general optimisation tool, capable of solving different problem variants and of including several objectives, the operations involved in the optimisation process do not consider any particular characteristics of the classical UC problem. Even so, when applied to instances with very particular structures, the computational results show the potential of this approach.

Abstract
One major practical problem when applying traditional metaheuristics seems to be their strong dependency on parameter tuning. This issue is frequently pointed out as a major shortcoming of metaheuristics and is often a reason for Decision-Makers to reject using this type of approach in practical situations. In this paper we present a new search strategy - Constraint Oriented Neighbourhoods - that tries to overcome the referred drawback. The aim is to control the grade of randomness of metaheuristics, by defining "special" neighbourhood movements, that lead to a more robust heuristic, less dependent on parameter tuning. This is achieved by selecting and applying particular movements that take into account the potential violation of problem constraints. The strategy is illustrated in a real problem arising in the area of Power Systems Management - the Unit Commitment Problem, the computational experiments on a set of problem instances systematically outperforming those presented in the literature, both in terms of efficiency, quality of the solution and robustness of the algorithm.

Abstract
Due to their efficiency and their interesting design and implementation features, metaheuristics have been used for a long time with success, in dealing with combinatorial problems. In recent years they have been applied to the Unit Commitment problem with rather interesting results that justify further research in the area. In this paper we present a Simulated Annealing approach to the Unit Commitment problem. Two coding schemes are compared, new neighbourhood structures are presented and some searching strategies are discussed. Preliminary computational experience, performed on some test instances, shows that this approach is flexible, effective and able to handle variations on the problem structure. © 2001 IEEE.