Abstract: The operating theater is one of the most expensive units in the hospital, representing up to 40% of the total expenses. Because of its importance, the operating room scheduling problem has been addressed from many different perspectives since the early 1960s. One of the main difficulties that
has reduced the applicability of the current results is the high variability in surgery duration, making schedule recommendations hard to implement.
In this work, we propose a time-indexed scheduling formulation to solve the operational problem. Our main contribution is that we propose the use of chance constraints related to the surgery duration's probability distribution for each surgeon to improve the scheduling performance. We show how to implement these chance constraints as linear ones in our time-indexed formulation, enhancing the performance of the resulting schedules significantly.
Through data analysis of real historical instances, we develop specific constraints that improve the schedule, reducing the need for overtime without affecting the utilization significantly. Furthermore, these constraints give the operating room manager the possibility of balancing overtime and utilization through a tunning parameter in our formulation. Finally, through simulations and the use of real instances, we report the performance for four different metrics, showing the importance of using historical data to get the right balance between the utilization and overtime.

Abstract: The MAXIMUM CAPTURE PROBLEM WITH RANDOM UTILITIES seeks to locate new facilities in a competitive market such that the captured demand of users is maximized, assuming that each individual chooses among all available facilities according to the well-know a random utility model namely the multinomial logit. The problem is complex mostly due to its integer nonlinear objective function. Currently, the most efficient approaches deal with this complexity by either using a nonlinear programing solver or reformulating the problem into a Mixed-Integer Linear Programing (MILP) model. In this paper, we show how the best MILP reformulation available in the literature can be strengthened by using tighter coefficients in some inequalities. We also introduce a new branch-and-bound algorithm based on a greedy approach for solving a relaxation of the original problem. Extensive computational experiments are presented, bench marking the proposed approach with other linear and non-linear relaxations of the problem. The computational experiments show that our proposed algorithm is competitive with all other methods as there is no method which outperforms the others in all instances. We also show a large-scale real instance of the problem, which comes from an application in park-and-ride facility location, where our proposed branch-and-bound algorithm was the most effective method for solving this type of problem. (C) 2015 Elsevier B.V. All rights reserved.

Abstract: There are significant research opportunities in the integration of Machine Learning (ML) methods and Combinatorial Optimization Problems (COPs). In this work, we focus on metaheuristics to solve COPs that have an important learning component. These algorithms must explore a solution space and learn from the information they obtain in order to find high-quality solutions. Among the metaheuristics, we study Hyper-Heuristics (HHs), algorithms that, given a number of low-level heuristics, iteratively select and apply heuristics to a solution. The HH we consider has a Markov model to produce sequences of low-level heuristics, which we combine with a Multi-Armed Bandit Problem (MAB)-based method to learn its parameters. This work proposes several improvements to the HH metaheuristic that yields a better learning for solving problem instances. Specifically, this is the first work in HHs to present Exponential Weights for Exploration and Exploitation (EXP3) as a learning method, an algorithm that is able to deal with adversarial settings. We also present a case study for the Vehicle Routing Problem with Time Windows (VRPTW), for which we include a list of low-level heuristics that have been proposed in the literature. We show that our algorithms can handle a large and diverse list of heuristics, illustrating that they can be easily configured to solve COPs of different nature. The computational results indicate that our algorithms are competitive methods for the VRPTW (2.16% gap on average with respect to the best known solutions), demonstrating the potential of these algorithms to solve COPs. Finally, we show how algorithms can even detect low-level heuristics that do not contribute to finding better solutions to the problem.& COPY

Abstract: We consider a family of competitive facility location problems in which a “newcomer” company enters the market and has to decide where to locate a set of new facilities so as to maximize its market share. The multinomial logit model is used to estimate the captured customer demand. We propose a first branch-and-cut approach for this family of difficult mixed-integer non-linear problems. Our approach combines two types of cutting planes that exploit particular properties of the objective function: the first one are the outer-approximation cuts and the second one are the submodular cuts. The approach is computationally evaluated on three datasets from the recent literature. The obtained results show that our new branch-and-cut drastically outperforms state-of-the-art exact approaches, both in terms of the computing times, and in terms of the number of instances solved to optimality. (C) 2017 Elsevier B.V. All rights reserved.

Abstract: We propose a novel two-phase bounding and decomposition approach to compute optimal and near-optimal solutions to large-scale mixed-integer investment planning problems that have to consider a large number of operating subproblems, each of which is a convex optimization. Our motivating application is the planning of power transmission and generation in which policy constraints are designed to incentivize high amounts of intermittent generation in electric power systems. The bounding phase exploits Jensen's inequality to define a lower bound, which we extend to stochastic programs that use expected-value constraints to enforce policy objectives. The decomposition phase, in which the bounds are tightened, improves upon the standard Benders' algorithm by accelerating the convergence of the bounds. The lower bound is tightened by using a Jensen's inequality-based approach to introduce an auxiliary lower bound into the Benders master problem. Upper bounds for both phases are computed using a sub-sampling approach executed on a parallel computer system. Numerical results show that only the bounding phase is necessary if loose optimality gaps are acceptable. However, the decomposition phase is required to attain optimality gaps. Use of both phases performs better, in terms of convergence speed, than attempting to solve the problem using just the bounding phase or regular Benders decomposition separately. (C) 2015 Elsevier B.V. and Association of European Operational Research Societies (EURO) within the International Federation of Operational Research Societies (IFORS). All rights reserved.

Abstract: Industrial manufacturing is often organized in assembly lines where a product is assembled on a se-quence of stations, each of which executes some of the assembly tasks. A line is balanced if the maximum total execution time of any station is minimal. Commonly, the task execution order is constrained by precedences, and task execution times are independent of the station performing the task. Here, we con -sider a recent variation, called the “(Calzedonia) Workload Allocation Problem” (WAP), where the prece-dences form a chain, and the execution time of a task depends on the worker executing it. This problem was recently proposed by Battarra et al. (2020) and it is a special case of the Assembly Line Worker As-signment and Balancing Problem Miralles et al. (2007) where precedence relations are arbitrary. In this paper we consider the computational complexity of the problem and prove its NP-hardness. To solve the problem, we provide different lower bounds and exact and heuristic procedures. The performance of the proposed methods is tested on previously proposed instances and on new, larger instances with the same characteristics. The results show that the proposed methods can solve instances with up to about 40 0 0 tasks and 29 workers, doubling the size of the instances that previously could be solved to optimality.

Abstract: Over the last decades, speculative investors in the FX market have profited in the well known currency carry trade strategy (CT). However, during currencies or global financial crashes, CT produces substantial losses. In this work we present a methodology that enhances CT performance significantly. For our final strategy, constructed backtests show that the mean-semivolatility ratio can be more than doubled with respect to benchmark CT. To do the latter, we first identify and classify CT returns according to their behavior in different regimes, using a Hidden Markov Model (HMM). The model helps to determine when to open and close positions, depending whether the regime is favorable to CT or not. Finally we employ a mean-semivariance allocation model to improve allocations when positions are opened. (C) 2016 Elsevier B.V. All rights reserved.

Abstract: Wildfires can pose a major threat to the secure operation of power networks. Chile, California, and Australia have suffered from recent wildfires that have induced considerable power supply cuts. Further, as power systems move to a significant integration of variable renewable energy sources, successfully managing the impact of wildfires on the power supply can become even more challenging due to the joint uncertainty in wildfire trajectories and the power injections from wind and solar farms. Motivated by this, this paper develops a practical decision-support approach that concatenates a stochastic wildfire simulation method with an attacker-defender model that aims to find a worst-case realization for (i) transmission line and generator contingencies, out of those that can potentially be affected by a given wildfire scenario, and for (ii) wind and solar power trajectories, based on a max-min structure where the inner min problem represents a best adaptive response on generator dispatch actions. Further, this paper proposes an evaluation framework to assess the power supply security of various power system topology configurations, under the assumption of limited transmission switching capabilities, and based on the simulation of several wildfire evolution scenarios. Extensive computational experiments are carried out on two representations of the Chilean power network with up to 278 buses, showing the practical effectiveness of the proposed approach for enhancing wildfire resilience in highly renewable power systems.

Abstract: One key assumption of Markowitz's model is that all traders act as price takers. In this paper, we extend this mean-variance approach in a setting where large investors can move prices. Instead of having an individual optimization problem, we find the investors' Nash equilibrium and redefine the efficient frontier in this new framework. We also develop a simplified application of the general model, with two assets and two investors to shed light on the potential strategic behavior of large and atomic investors. Our findings validate the claim that large investors enhance their portfolio performance in relation to perfect market conditions. Besides, we show under which conditions atomic investors can benefit in relation to the standard setting, even if they have not total influence on their eventual performance. The 'two investors-two assets' setting allows us to quantify performance and do sensitivity analysis regarding investors' market power, risk tolerance and price elasticity of demand. Finally, for a group of well known ETFs, we empirically show how price variations change depending on the volume traded. We also explain how to set up and use our model with real market data. (C) 2016 Elsevier B.V. All rights reserved.