Conference Chairs

List of Chairs are available in this pdf.


Conference Schedule
Talk: Advances in Evolutionary Multitasking for Vehicle Routing Evolutionary multi-tasking (EMT) is a recently emerged search paradigm in the realm of evolutionary computation. In contrast to traditional evolutionary algorithm (EA) which solves a single task in a single run, the EMT conducts evolutionary search concurrently on multiple search spaces corresponding to different tasks or optimization problems, each possessing a unique function landscape. By leveraging the useful knowledge across tasks while the evolutionary search progresses online, the EMT has demonstrated superior optimization performance in terms of search efficiency and effectiveness on continuous as well as discrete optimization problems, when compared to its single-task counterparts. Vehicle routing problem (VRP) is a generalization of the Traveling Salesman Problem (TSP), which has been proved to be NP-hard. To solve the VRP, traditional single-task optimization methods ignored the similarity shared among VRPs, which could lead to enhanced vehicle routing process when properly harnessed. Taking this cue, several EMT algorithms have been proposed towards an enhanced vehicle routing process in the literature. In this talk, I will present our recent progresses on advanced EMT algorithm designs for VRPs, such as Capacitated VRP and VRP with heterogeneous capacity, time window as well as occasional drivers. Some key issues of EMT algorithm designs for VRPs will also be discussed.

Lixin Tang, Northeastern University, China
Lixin Tang is a Chair Professor of the Institute of Industrial & Systems Engineering of Northeastern University, the Director of Key Laboratory of Data Analytics and Optimization for Smart Industry under Ministry of Education of China, and the Head of Centre for Artificial Intelligence and Data Science. His research interests cover plant-wide production and inventory planning, production batching and scheduling, data analytics and optimization for production process, integer optimization, computational intelligent optimization and machine learning, dynamic optimization and reinforcement learning, and engineering applications in manufacturing (steel, petroleum-chemical, nonferrous), energy, resources industry and logistics systems.

Talk: Data Analytics and Optimization for Steel Industry
Some interesting topics on the production, logistics and energy analytics and optimization in steel industry will be discussed in this talk, including: 1) production scheduling in steel-making and hot/cold rolling operations; 2) logistics scheduling in storage/stowage, shuffling, transportation and (un)loading operations; 3) energy optimization including energy allocation and coordinated planning and scheduling of production and energy; 4) data analytics including dynamic analytics for BOF steelmaking process based on multi-stage modeling; temperature prediction for blast furnace; temperature prediction for molten iron in transportation process; analytics for description, diagnostics, and prediction of energy generation and consumption; temperature prediction for reheat furnace based on mechanism and data; strip quality analytics for continuous annealing based on multi-objective ensemble learning; process monitoring and diagnosis for continuous annealing based on mechanism and data.  (1999)(2000), and as Treasurer of INFORMS (2011INFORMS ( -2014. He served on the State of Ohio Steel Industry Advisory Council (1997)(1998)(1999)(2000)(2001)(2002). He has been a visiting professor at the Wharton School (University of Pennsylvania) and Kellogg School (Northwestern University).
He is the owner of a consulting business, CDOR, which provides business solutions to the Ohio business and government communities, and advice on intellectual property issues to New York City law firms. In 2018, he served as the 24th President of INFORMS, and introduced an outreach program to leading policymakers at the White House and on Capitol Hill.

Talk: A New Design for Tournaments
The tournament designs that are used for numerous professional and amateur sports events all over the world suffer from a number of well documented deficiencies. These deficiencies typically reduce the fairness, excitement value, profitability, and credibility of those tournaments. We describe a new design that substantially reduces these problems. We validate our design using both data from professional sports tournaments and an extensive sensitivity analysis with simulated data. Cutting and Packing problems are hard combinatorial optimization problems that arise under several practical contexts, whenever big pieces of raw-material have to be cut into smaller items, or small items have to be packed inside a larger container, so that waste is minimized. These problems include hard geometric constraints when dealing with the optimization layer. I have also worked on Vehicle Routing Problem. My research on Lotsizing and Scheduling problems in industrial contexts mainly builds on my expertise on Metaheuristics.
More recently I have worked on the use of the quantitative methods, provided by Operations Research and Management Science, to support decision making in Higher Education institutions management, which includes workload models, sustainability, institutional benchmarking and assessment and evaluation of institutions and teaching staff.
Topic: Tackling uncertainty in combinatorial optimization problems: using metaheuristics to efficiently co-generate scenarios and solutions Uncertainty is receiving increasing attention, in the past years, from the Operational Research community. Methods that acknowledge uncertainty and incompleteness of information are an important research trend. Scenarios arise as key components in many of these methods, as instruments to deal with uncertainty. However, the scenario generation process is often unrealistically simplified. We propose that metaheuristics, namely based on genetic algorithms, can generate relevant and complex scenarios, without requiring a priori probability distributions. This is of particular interest in practical applications where there are many uncertain parameters, and it is significantly difficult to define their characteristics accurately. To address two-stage stochastic problems, we propose a method based on a co-evolutionary metaheuristic, where solutions and scenarios are generated and evolve in parallel. The goal of the evolution of the solution population is to obtain values for the first-stage decisions that perform well when compared with the scenario population. The goal of the evolution of the scenario population is to diversify the impact of its elements on the value of solutions. This methodology is able to support decision-makers with different risk profiles. To illustrate the method, we apply it to the integrated problem of fleet management and pricing for car rental companies under demand and competitor pricing uncertainty. When planning a selling season, a car rental company must decide on the number and type of vehicles in the fleet to meet demand. The demand for the rental products is uncertain and highly pricesensitive, and thus capacity and pricing decisions are interconnected. Moreover, since the products are rentals, capacity "returns". This creates a link between capacity with fleet deployment and other tools that allow the company to meet demand, such as upgrades, transferring vehicles between locations or temporarily leasing additional vehicles. An ongoing extension of this work is an innovative scenario generator, based on this idea of impact diversity, that quickly provides representative sets of scenarios that can be used to feed not only a genetic algorithm but any stochastic solution method.

Ling Wang, Tsinghua Univ, China
Ling Wang received the B.Sc. and Ph.D. degrees from Tsinghua University, Beijing, China, in 1995 and1999, respectively, and now is a tenured Professor in Tsinghua Univ. His research interests mainly include intelligent optimization, scheduling and applications. He has authored 5 academic books and more than 150 SCI-indexed papers.

Data-Driven Intelligent Optimization Scheduling
Optimization and scheduling problems are the significant issues faced by the manufacturing industry. Over the past few decades, an impressive number of data-driven intelligence algorithms were reported for solving the engineering optimization and scheduling problems (EOSPs). This talk first shows the complexities of the EOSPs; and then introduces a unified framework for the population-base intelligent optimization techniques from a systematic perspective as well as an integrated intelligent optimization framework; finally presents some typical research work in terms of theoretical analysis, constrained optimization and intelligent scheduling algorithms. The primary aim of this talk is to show that intelligent algorithms are powerful and general solution tools for solving the EOSPs, while it is more important to incorporate the problem-specific knowledge into the algorithms for solving specific problems.

Tutorial speakers:
Tutorial 1: Appointment Scheduling Speaker: Guohua Wan, Department of Management Science, Shanghai Jiaotong University Abstract: Appointment scheduling is concerned with allocation of service times to jobs so as to optimize certain performance measures in service systems. Appointment scheduling systems are widely used to manage access to services. In this tutorial, we will summarize the research development of models and algorithms in appointment scheduling, in particular the models and algorithms in healthcare services, and discuss future research opportunities in appointment scheduling. Abstract: Multitasking, defined as "the performance of multiple tasks at one time," is a behavioral phenomenon that can be observed frequently in our daily life. This behavioral phenomenon can result in a significant loss of productivity. Principal motivations for multitasking in administrative and business processes include (i) a need to feel or appear productive, (ii) a need to demonstrate progress on different tasks or treat task owners equitably, (iii) anxiety about the processing requirements of waiting tasks, (iv) a need for variety in work, and (v) interruption by routine scheduled activities. This tutorial discusses various scheduling models that capture these multitasking motivations. Unlike classical scheduling models where no work is ever processed except as a result of a deliberate decision by a decision maker, our scheduling models assume that multitasking is inadvertent. For these new scheduling models, we discuss their problem variants, computational complexities, optimality properties, solution methods, and possible extensions. Abstract: Cooperative game theory focuses on schemes that lead to a global collaboration among multiple independent decision makers. In cooperative game theory, one basic concept is the allocation in the core that characterizes how the players shall share the cost/benefit in a way acceptable to all sub-coalitions. Unfortunately, it is well known that many cooperative games have an empty core, including games concerning scheduling problems. For such games the global collaboration will not be sustainable. We consider a situation where an outside party has the need to stabilize the ground coalition because, for example, the best social welfare can be achieved only when all players collaborate. We introduce a few economic treatments that can be used by the outside party such as providing subsidy and charging penalty. These treatments, including their concepts and implementations, are demonstrated by games related to scheduling problems. The vehicle scheduling problem (VSP) is concerned with determining the most efficient allocation of vehicles to carry out all the trips in a given timetable. The duration of each trip (called trip time) is normally assumed to be fixed. However, in practice, the trip times vary due to the variability of traffic, driving condition and passengers' behavior. The compiled schedule to some extent is hence hard to be adhered to. This paper proposes a new VSP model based on variable trip times. Instead of being a fixed value, the duration of a trip falls into a time range, which can be easily set based on the schedulers' experience or the data collected by automatic vehicle location (AVL) systems. Within this range, an expected trip time is provided according to the schedulers' expectation of on-time performance. This new model can reduce schedulers' pressure on setting fixed scheduled trip times. Moreover, computational results generated using CPLEX show that this model can increase the on-time performance of resulting schedules without increasing the fleet size. Abstract: With the advance of deep learning, deep reinforcement learning have become quite popular in AI research community and industry as well. Deep reinforcement learning is viewed as one of the most promising way to achieve artificial general intelligence. Recent years have witnessed a wide range of DRL applications in robotics, e-commerce, and games and so on. Compared with single-agent learning, concurrent learning among multiple agents (aka multi-agent learning) are more common in our human society and human-made systems and have received more attention in the last few years. In this talk, I will first introduce the background of multi-agent reinforcement learning, and then talk about recent research progresses in deep multi-agent reinforcement learning and its applications in industries. Abstract: This study addresses a vehicle routing problem (VRP) in which demands are discrete, split delivery is allowed, service time is proportional to the units of delivered products, multiple time windows are provided and the demand of each customer must be delivered in only one time window (this requirement is termed synchronization constraint). We formulate this problem into an arc-flow model and a set-covering model. Then, we propose a branch-and-price-and-cut algorithm to solve the problem. We compare our branch-and-price-and-cut algorithm with CPLEX based on 252 randomly generated instances and the computational results demonstrate the effectiveness of our proposed algorithm.

Paper list
Paper ID Authors and title How to Get to Ningbo?

Plane
Many people fly to Ningbo-Lishe International Airport (NGB). Several major airlines offer flights from the US and Europe with ongoing connections to Ningbo from larger Chinese cities like Shanghai, Hangzhou, Hong Kong, Guangzhou and Beijing. The airport is eight miles southwest of the city center, and there is a regular shuttle bus service from the arrivals terminal to South Ningbo Railway station for ¥10. Taxis are also available and cost ¥50.

Train
Ningbo can be reached by train from most large cities and it's a relatively inexpensive option. A first-class return from Shanghai costs ¥450 and from Hangzhou, a first-class return is ¥250. Roombooking can be made by calling the above telephone number. Please do mention "MISTA 2019" in your phone call.