On behalf of the Editorial Board, I would like to thank the hundreds of people who acted as reviewers for the INFORMS Journal on Computing (IJOC) during the past year. Reviewers are the cornerstone of the peer review system and give their time and expertise unselfishly. This year, many went beyond the call of duty and submitted reviews that warranted meritorious recognition. These reviews were especially insightful and detailed. They were nominated by our editorial team as exemplifying the highest standards of peer review. We also list every reviewer who completed a review. IJOC reviewers, please know you are greatly appreciated!

Truck-drone delivery systems have been proposed for sustainable and economical last-mile distribution, especially in urban environments. To widen the service range, some works have recommended adding facilities such as drone stations, considering the problem in discrete space by choosing from a pre-defined set. In this paper, an evolutionary optimization approach to the design decision of where to locate drone stations in the continuous plane is introduced, modeled, and solved. Drone stations serve as facilities for storage, charging, and launching. A truck (or other land transport means) transports parcels to the drone stations from a depot and the drones launch from the stations and deliver the parcels to each customer. The objective is to determine the positions of the drone stations in two-dimensional space and establish the shortest fixed truck route from the depot through all the stations and returning to the depot. The problem is constrained by the radius of service for each drone and all customers must be served, if possible. We formulate the problem as a constrained nonlinear optimization problem and present two versions of an algorithm using particle swarm optimization with a subordinate dynamic program. Computational results show that our approach achieves much better results than a standard commercial nonlinear solver in a similar amount of computational time for both maximizing coverage of customers and minimizing distance of the truck delivery route. A design case study concerning healthcare delivery throughout the Birmingham, Alabama (USA) metropolitan area is provided.

Public sentiment can impact the implementation of public policies and even cause policy failure if public support is not received. Therefore, knowledge of public sentiment concerning new and emerging policies is critical for policymakers. During the coronavirus disease 2019 (COVID-19) pandemic, several precautionary measures have been suggested in an attempt to delay or mitigate the spread of the virus. This study presents a framework that applies natural language processing (NLP) techniques, such as sentiment and bigram analyses, to characterize the public sentiment on three prominent mitigation measures (mask wearing, social distancing, and quarantine) as shared by Twitter users in the United States. As part of the framework, we apply a bigram graph-based approach to visualize the most frequent topics in Twitter discussions during the COVID-19 pandemic. The objective is to provide insights into the most commonly discussed topics among Twitter users with similar demographic characteristics (e.g., age and gender). The sentiment and bigram analyses identified the most frequently discussed topics expressing both positive and negative sentiments among different age and gender groups. Discussions containing positive sentiment prevailed and revolved around the benefits of the measures and trust in the government, while the topics of negative sentiment involved conspiracy theories, skepticism, and distrust of government mandates. It is also notable that the discussions among people 19–29 and over 40 years old focus on government officials and political parties, benefits or inefficiency of mitigation measures, and conspiracy theories more often than other demographic groups. Our proposed approaches and results offer a novel and potentially valuable contribution to public policymakers.

Recent developments in computational science and engineering allow a great deal of experimental work to be conducted through computer simulation. In a simulation experiment, a model of the phenomena to be studied is run in a computing environment under varying model and environment settings. As models are adjusted to experimental procedures and execution environments, variations arise. Models also evolve in time. Thus, models must be managed. We propose to bring Global Model Management (GMM) to bear on simulation experiment management by using techniques and tools from megamodeling. The proposed approach will facilitate model management tasks by providing an interface to query the model repository, relate models with each other, and apply model transformations from/to simulation models. Our proposed Megamodel for Simulation Experiments is based on SED-ML (Simulation Experiment Description Markup Language).

In telecommunication network design problems, survivability and reliability are often used to evaluate quality of service while usually ignoring link capacity. In this paper, a new metric that combines network reliability with network resilience is presented to measure reliability/survivability effectively for capacitated networks. Capacitated resilience is compared with well-known network reliability/survivability metrics (k-terminal reliability, all-terminal reliability, traffic efficiency, and k-connectivity), and its benefits and computational efficiency are discussed. An application is shown using heterogeneous wireless networks (HetNets). With the growing use of new telecommunication technologies such as 4G and wireless hotspots, HetNets are gaining more attention. The source of heterogeneity of a HetNet can either be the differences in nodes (such as transmission ranges, failure rates, and energy levels) or the differences in services offered in the network (such as GSM and WiFi).

Fuzzy rule based systems have been successfully applied to the pattern classification problem. In this research, we proposed an iterative mixed-integer programming algorithm to generate fuzzy rules for fuzzy rule-based classification systems. The proposed model is capable of assigning the attributes to the antecedents of rules so that their inclusion enhances the accuracy and coverage of that rule. To generate several diverse rules per class, the integer programming model is run iteratively and all samples predicted correctly are temporarily removed from the training dataset in each iteration. This process ensures that subsequent rule covers new samples in the associated class. The proposed model was evaluated on the benchmark datasets from the UCI repository and this comparative study verifies that this approach extracts accurate rules and has advantage over conventional approaches for high dimensional datasets.

Facility layout problems have drawn much attention over the years, as evidenced by many different versions and formulations in the manufacturing context. This paper is motivated by semiconductor manufacturing, where the floor space is highly expensive (such as in a cleanroom environment) but there is also considerable material handling amongst machines. This is an integrated optimization task that considers both material movement and manufacturing area. Specifically, a new approach combining multiobjective tabu search with linear programming is proposed for an extended double row layout problem, in which the objective is to determine exact locations of machines in both rows to minimize material handling cost and layout area where material flows are asymmetric. First, a formulation of this layout problem is established. Second, an optimization framework is proposed that utilizes multiobjective tabu search and linear programming to determine a set of non-dominated solutions, which includes both sequences and positions of machines. This framework is applied to various manufacturing situations, and compared with an exact approach and a popular multiobjective genetic algorithm optimization algorithm. Experimental results show that the proposed approach is able to obtain sets of Pareto solutions that are far better than those obtained by the alternative approaches.

This paper addresses subsonic airfoil optimization using Evolution Strategies (ES) and devises a means of defining the airfoil geometry that reduces unnecessary restrictions in the search space. The solution encoding uses Bezier Control Points to define the geometry of the airfoil, but does not restrict the movement of the control points as was common in previous airfoil optimization algorithms. The ES move operator combined with this improved solution encoding expands the search space to include superior solutions while also enabling a more efficient search to reduce computational cost.

We present an application of the Single Minute Exchange of Dies (SMED) methodology used in lean automotive manufacturing for the design and implementation of metaheuristic algorithms. This methodology allows the design of algorithm implementations that are flexible and enables algorithm designers to modify the algorithm's configuration quickly and with a minimum amount of errors. This comes with little downside as measured by computational effort and engenders a wider variety of experimental scenarios to be tested.

We consider various bi-objective models for the semi desirable facility location problem. In these problems, the disservice caused by the facility is traditionally measured by distance-related objective functions. In this paper, we modify the objective function representing the disservice using the Lorenz curve and the Gini coefficient. Both of these concepts are widely used in the economics literature to measure the discrepancy in wealth distribution within a population. The use of the Gini coefficient enables the measurement of how the disservice caused by the facility varies across different Pareto optimal solutions. We use a bi-objective particle swarm optimizer (bi-PSO) to compare how the change in the objective function representing the disservice affects the recommended location of the facility. Results suggest that some solutions identified as "Pareto optimal" by traditional formulations are dominated by other solutions when the Gini coefficient is used. Additionally, the use of the Gini coefficient causes a change in the "optimal" location of a semi desirable facility in some instances.

This paper presents a biobjective genetic algorithm (GA) to design reliable two-node connected telecommunication networks. Because the exact calculation of the reliability of a network is NP-hard, network designers have been reluctant to use network reliability as a design criterion; however, it is clearly an important aspect. Herein, three methods of reliability assessment are developed: an exact reliability calculation method using factoring, an efficient Monte Carlo estimation procedure using the sequential construction technique and network reductions, and an upper bound for the all-terminal reliability of networks with arbitrary arc reliabilities. These three methods of reliability assessment are used collectively in a biobjective GA with specialized mutation operators that perturb solutions without disturbing two-node connectivity. Computational experiments show that the proposed approach is tractable and significantly improves upon the results found by single-objective heuristics.

An agent location optimization model for military mobile ad-hoc networks is described. A mobile ad-hoc network (MANET) is a self-configuring network of autonomous agents designed to continuously support users who change the topology of the network dynamically and independently. The autonomous agents are controlled to maximize the connectivity of user nodes that move within the target region considering tactical military aspects. The primary objective of the agent location model is to maximize the connectivity and quality of communication between user nodes and a control node. To support military applications, a new approach, the Pre-deployed Agent Level (PAL) is introduced and a particle Swarm Optimization (PSO)-based heuristic with PAL is developed to solve the problem under three user mobility models. The focus of the paper is to determine the effect of PAL on the communication quality in the network.

This paper describes a case study of the development and testing of a prototype system to support condition-based maintenance of the door systems of airport transportation vehicles. Every door open/close cycle produces a ¿signature¿ that can indicate the current degradation level of the door system. A combined statistical and neural network approach was used. Time, electrical current and voltage signals from the open/close cycles are processed in real-time to estimate, using the neural network, the condition of the door set relative to maintenance needs. Data collection hardware for the vehicle was designed, developed and tested to monitor door characteristics to quickly predict degraded performance, and to anticipate failures. The prototype system was installed on vehicle door sets at the Pittsburgh International Airport and tested for several months under actual operating conditions.

This paper puts forth a new encoding method for using neural network models to estimate the reliability of telecommunications networks with identical link reliabilities. Neural estimation is computationally speedy, and can be used during network design optimization by an iterative algorithm such as tabu search, or simulated annealing. Two significant drawbacks of previous approaches to using neural networks to model system reliability are the long vector length of the inputs required to represent the network link architecture, and the specificity of the neural network model to a certain system size. Our encoding method overcomes both of these drawbacks with a compact, general set of inputs that adequately describe the likely network reliability. We computationally demonstrate both the precision of the neural network estimate of reliability, and the ability of the neural network model to generalize to a variety of network sizes, including application to three actual large scale communications networks.

In this study, a novel approach is used to identify nondominated solutions to multiobjective optimization problems. The method is composed of a Particle Swarm Optimizer (PSO) coupled with a neural network. The PSO is used to find an initial set of nondominated solutions. These nondominated solutions are then used to construct a general regression neural network that generates a considerably larger set of nondominated solutions. Our neural network enhancement process is demonstrated on a test suite of six instances of bi-criteria semidesirable facility location problems. Results show that the set of nondominated solutions developed by the neural network is, on average, 25 times larger than the initial set found by PSO, and in many instances dominate those identified by PSO. The method developed within is straightforward and general and is a new alternative to multiobjective optimization with decision variables in continuous space.

As high-speed networks have proliferated across the globe, their topologies have become sparser due to the increased capacity of communication media and cost considerations. Reliability has been a traditional goal within network design optimization of sparse networks. This paper proposes a genetic approach that uses network resilience as a design criterion in order to ensure the integrity of network services in the event of component failures. Network resilience measures have been previously overlooked as a network design objective in an optimization framework because of their computational complexity - requiring estimation by simulation. This paper analyzes the effect of noise in the simulation estimator used to evaluate network resilience on the performance of the proposed optimization approach.

A non-deterministic decoding algorithm for binary coded genetic algorithms is presented. The proposed algorithm enhances the precision of the GA solutions by introducing a Gaussian perturbation to the decoding function. This non-deterministic decoding enables individuals to represent any point in the continuum instead of finite discrete points. As the generations evolve, information gathered from the most fit members is continuously used to rearrange the binary representation grid on the search space, thus establishing a search memory such that the best known individual is always positioned at the center of the Gaussian offset.

The problem of locating a semi-desirable facility on a plane is considered. A bi-criteria model is used. One of the criteria is well known minimum criterion. The second criterion is a weighted sum of Euclidean distances raised to the power of negative one. This function represents the aggregate undesirable effects of the facility and it is also a minimization problem. The bi-criteria model consists of a linear combination of these two criteria. The proposed model solved by particle swarm optimization provides better results than a previously proposed heuristic.

This paper puts forth a general method to optimize constrained problems effectively when using tabu search. An adaptive penalty approach is used that exploits the short-term memory structure of the tabu list along with the long-term memory of the search results. It is shown to be effective on a variety of combinatorial problems with different degrees and numbers of constraints. The approach requires few parameters, is robust to their setting, and encourages search in promising regions of the feasible and infeasible regions before converging to a final feasible solution. The method is tested on three diverse NP-hard problems, facility layout, system reliability optimization, and orienteering, and is compared with two other penalty approaches developed explicitly for tabu search. The proposed memory-based approach shows consistent strong performance.

This paper uses an ant colony meta-heuristic optimization method to solve the redundancy allocation problem (RAP). The RAP is a well known NP-hard problem which has been the subject of much prior work, generally in a restricted form where each subsystem must consist of identical components in parallel to make computations tractable. Meta-heuristic methods overcome this limitation, and offer a practical way to solve large instances of the relaxed RAP where different components can be placed in parallel. The ant colony method has not yet been used in reliability design, yet it is a method that is expressly designed for combinatorial problems with a neighborhood structure, as in the case of the RAP. An ant colony optimization algorithm for the RAP is devised & tested on a well-known suite of problems from the literature. It is shown that the ant colony method performs with little variability over problem instance or random number seed. It is competitive with the best-known heuristics for redundancy allocation.

The exact calculation of all-terminal network reliability is an NP-hard problem, with computational effort growing exponentially with the number of nodes and links in a network. Because of the impracticality of calculating all terminal network reliability for networks of moderate to large size, Monte Carlo simulation methods to estimate network reliability and upper and lower bounds to bound reliability have been used as alternatives. In this study, an artificial neural network (ANN) is used to estimate all-terminal network reliability for networks with both homogeneous link reliability. Two forms of design methods for generating training data sets for networks with homogeneous and heterogeneous link reliability are compared. These experimental design and randomized design.

This paper presents two meta-heuristic techniques, ant colony optimization and tabu search, for the orienteering problem, a general version of the well-known traveling salesman problem with many relevant applications in industry. Both algorithms are compared to other heuristics in the literature. Results on 67 test problems show that the ant colony optimization method and tabu search method perform as well, or better, in all cases and do so at competitive computational cost.

This paper formulates and compares four new approaches to optimally locate the input and output station for each department within a facility design such that material handling costs are minimized. This problem is an NP-hard combinatorial problem with many real-life applications of considerable economic consequence. A genetic algorithm (GA) is shown to be an effective and efficient optimization method when compared to integer programming, simulated annealing, and three versions of a greedy constructive heuristic on a suite of test problems of varying size. Seeding versus random initialization of GA populations are compared.

This paper proposes an evolutionary approach for constructing 2-edge-connected minimal Steiner graphs spanning n points given in the rectilinear plane. The 1-edge-connected version of this problem is known as the rectilinear Steiner tree problem and has been widely studied. Despite the possible application areas, the 2-edge-connected problem has not received the same attention. In this paper, some properties of an optimal solution to the problem are used to develop an encoding scheme. The proposed evolutionary approach is compared on a test problem with the optimal TSP tour of the given points.

This paper presents a genetic algorithm to solve the orienteering problem, which is concerned with finding a path between a given set of control points, among which a start and an end point are specified, so as to maximize the total score collected subject to a prescribed time constraint. Employing three sets of test problems from the literature, the performance of the genetic algorithm is evaluated against problem specific heuristics and an artificial neural network.

The paper presents a hybrid approach of a genetic algorithm (GA) and local search algorithms for the backbone design of communication networks. The backbone network design problem is defined as finding the network topology minimizing the design/operating cost of a network under performance and survivability considerations. This problem is known to be NP-hard. In the hybrid approach, the local search algorithm efficiently improves the solutions in the population by using domain-specific information while the GA recombines good solutions in order to investigate different regions of the solution space. The results of the test problems show that the hybrid methodology improves upon previous approaches.

The paper solves the redundancy allocation problem of a series-parallel system by developing and demonstrating a problem-specific ant system. The problem is to select components and redundancy levels to maximize system reliability, given system-level constraints on cost and weight. The ant system algorithm presented in the paper is combined with an adaptive penalty method to deal with the highly constrained problem. An elitist strategy and mutation are introduced to our AS algorithm. The elitist strategy enhances the magnitude of the trails of good selections of components. The mutated ants can help explore new search areas. Experiments were conducted on a well known set of sample problems proposed by D.E. Fyffe et al. (1968).

The paper discusses the first-hand experiences of the author in developing and teaching a graduate level engineering course in adaptive optimization methods inspired by nature. The paper discusses course content, textbooks and supplementary written material, software and computer projects, and grading and evaluation. This course has encouraged many students to pursue research in evolutionary computation, tabu search or simulated annealing, however it is continually being modified to reflect the many changes occurring in the field.

The potential use of backpropagation networks, cascade-correlation learning networks, and radial basis function networks in developing metamodels to assist in performing sensitivity analysis of capital investment decisions is examined. The neural network metamodel approach is illustrated through a case study. It is shown that the performance of backpropagation and cascade-correlation learning metamodels is comparable with the traditional polynomial regression metamodel.

Investigates hybrid approaches of statistical resampling methodologies to neural network validation when the data is sparse. Specifically, cross validation, grouped cross validation, bootstrapping and resubstitution are considered. Resampling methods have been used occasionally in the literature for neural network validation when the amount of data is limited. This research paper uses simulated data to examine different resampling estimates first, then applies stepwise regression to identify different hybrid estimates. The best hybrid estimate, which combines the resampling estimates obtained from twenty bootstrapped samples, five bootstrapped samples and two-fold cross validation, is nearly unbiased with less variability. This hybrid requires constructing twenty-seven validation networks.

This paper presents a meta-heuristic approach using genetic algorithm (GA) and cost consideration to optimize the reliability of computer communication networks. When a network topology is known, the problem of choosing the types of links and computer systems among alternatives which have different system reliability and costs is an NP-hard combinatorial problem. If there are m alternative links and k alternative computer systems, the search space for a known network topology with |L| links and |N| nodes is m/sup |L|/. k/sup |N|/. The heuristic is shown to be effective and computationally efficient compared to optimal solutions on a set of test problems.

The exact calculation of all-terminal network reliability is an NP-hard problem, with computational effort growing exponentially with the number of nodes and links in the network. Due to the impracticality of calculating all-terminal network reliability for networks of moderate to large size, Monte Carlo simulation methods have been used to estimate the network reliability and reliability upper and lower bounds. This paper puts forth an alternative to the estimation of all-terminal network reliability by using artificial neural network predictive models. Neural networks are constructed, trained and validated using alternative network topologies, a network reliability upper bound and the exact network reliability as a target. A hierarchical approach is used: a general neural network screens all network designs for reliability followed by a specialized neural network for highly reliable network designs. Results on a ten node problem are given using a grouped cross validation approach.

Neural networks must be constructed and validated with strong empirical dependence, which is difficult under conditions of sparse data. The paper examines the most common methods of neural network validation along with several general validation methods from the statistical resampling literature, as applied to function approximation networks with small sample sizes. It is shown that an increase in computation, necessary for the statistical resampling methods, produces networks that perform better than those constructed in the traditional manner. The statistical resampling methods also result in lower variance of validation, however some of the methods are biased in estimating network error.

This paper presents a genetic algorithm (GA) with specialized encoding, initialization, and local search operators to optimize the design of communication network topologies. This NP-hard problem is often highly constrained so that random initialization and standard genetic operators usually generate infeasible networks. Another complication is that the fitness function involves calculating the all-terminal reliability of the network, which is a computationally expensive calculation. Therefore, it is imperative that the search balances the need to thoroughly explore the boundary between feasible and infeasible networks, along with calculating fitness on only the most promising candidate networks. The algorithm results are compared to optimum results found by branch and bound and also to GA results without local search operators on a suite of 79 test problems. This strategy of employing bounds, simple heuristic checks, and problem-specific repair and local search operators can be used on other highly constrained combinatorial applications where numerous fitness calculations are prohibitive.

The application of genetic algorithms (GA) to constrained optimization problems has been hindered by the inefficiencies of reproduction and mutation when feasibility of generated solutions is impossible to guarantee and feasible solutions are very difficult to find. Although several authors have suggested the use of both static and dynamic penalty functions for genetic search, this paper presents a general adaptive penalty technique which makes use of feedback obtained during the search along with a dynamic distance metric. The effectiveness of this method is illustrated on two diverse combinatorial applications: (1) the unequal-area, shape-constrained facility layout problem and (2) the series-parallel redundancy allocation problem to maximize system reliability given cost and weight constraints. The adaptive penalty function is shown to be robust with regard to random number seed, parameter settings, number and degree of constraints, and problem instance.

A problem-specific genetic algorithm (GA) is developed and demonstrated to analyze series-parallel systems and to determine the optimal design configuration when there are multiple component choices available for each of several k-out-of-n:G subsystems. The problem is to select components and redundancy-levels to optimize some objective function, given system-level constraints on reliability, cost, and/or weight. Previous formulations of the problem have implicit restrictions concerning the type of redundancy allowed, the number of available component choices, and whether mixing of components is allowed. GA is a robust evolutionary optimization search technique with very few restrictions concerning the type or size of the design problem. The solution approach was to solve the dual of a nonlinear optimization problem by using a dynamic penalty function. GA performs very well on two types of problems: (1) redundancy allocation originally proposed by Fyffe, Hines, Lee, and (2) randomly generated problem with more complex k-out-of-n:G configurations.