Asia-Pacific Journal of Operational Research

This paper develops the game models of two symmetric supply chains, each consisting of one manufacturer and one retailer, while both retailers compete in the market with a linear function. The disclosure mechanism is designed when the information of the disrupted demand is asymmetric between supply side and retail side. We first study the model with the full information as a benchmark to explore the effect of asymmetric information on the system. In the case, each manufacturer maximizes her profit while the downstream retailer only obtains the reservation profit. For the case of asymmetric information, each manufacturer can obtain the real information of the disrupted demand by using a menu of contract bundles. For each information structure, there are always robust regions for each manufacturer’s original trading quantity scheme. That is, when the disrupted amount of the demand is sufficiently small, the trading quantity will be unchanged. However, some special measures, e.g., the higher unit wholesale price, should be taken to prevent the retailer from deviating the trading quantity scheme. The high-disruption retailer gets the higher profit due to the information rent. Compared with a single supply chain, Cournot competition results in the less retail price and the lower performance for the whole system.

Scholars from Mainland China have published 8541 papers in 79 Operations Research and Management Science (OR/MS) journals available on the ISI Web of Science (WOS). Data from 2001 to 2013 constitute the basis of the bibliometric method used in this paper for statistical analysis of status of OR/MS research in Mainland China, from the perspectives of number of papers, quality of papers, most productive institutions, co-author analysis, citations analysis and research topic analysis, etc. The results show that although the number of OR/MS papers by Mainland China scholars grew exponentially during this period, the quality of papers needs to be further improved. Besides, the competition among the institutions in Mainland China is very fierce. Furthermore, the co-author analysis shows that the depth of cooperation has increased steadily and yet the breadth of cooperation has much room for development. Moreover, from the perspective of citations, only a small number of papers by Mainland China scholars have influenced the international OR/MS field. In addition, analysis of research topics shows that Mainland China scholars have been studying the OR/MS problems from a variety of perspectives. Additionally, we also visualize the relationship of the keywords to describe the status of development of the OR/MS discipline in Mainland China, and conclude that the connection nodes will have greater potential in the future.

As a basic part of organizations’ logistics management, purchasing function has supplier selection as one of its main responsibilities. One of the main objectives a buyer follows in supplier selection is to determine optimal quota to be allocated to each supplier. How to allocate orders to different suppliers is as important task as it may affect efficiency of the whole chain. Also, due to variations in real-world business environment, order allocation process is always associated with uncertainties that make it complicated. Therefore, a three-stage integrated framework with environmental uncertainties considered is proposed to allocate orders; this framework can determine qualified suppliers to whom it assigns optimal quota. Considering multi-period purchases and uncertainties, this framework presents a multi-objective nonlinear programming model to determine optimal quota to be allocated to each qualified supplier within each specified period. In order to have the order allocation process closer to real-world cases while increasing the reliability of the obtained solutions, time value of money, inflation, transportation modes, supplier’s profit, and pricing strategy are considered in this model. According to uncertain structure of the proposed model, a solution strategy is proposed to convert this model into a single-objective deterministic model. Then, the resulted single-objective deterministic model is solved by proposing three evolutionary metaheuristic algorithms based on cuckoo optimization algorithm and imperialist competitive algorithm. Finally, a sample problem is presented together with some statistical tests and sensitivity analyses to assess and examine the proposed framework.

In this paper, a kernel-free soft quadratic surface support vector machine model is proposed for binary classification directly using a quadratic function for separation. Properties (including the solvability, uniqueness and support vector representation of the optimal solution) of the proposed model are derived. Results of computational experiments on some artificial and real-world classifying data sets indicate that the proposed soft quadratic surface support vector machine model may outperform Dagher’s quadratic model and other soft support vector machine models with a Quadratic or Gaussian kernel in terms of the classification accuracy and robustness.

The aim of this paper is to develop an effective method for solving matrix game with payoffs of trapezoidal fuzzy numbers (TrFNs). The method always assures that players’ gain-floor and loss-ceiling have a common TrFN-type fuzzy value and hereby any matrix game with payoffs of TrFNs has a TrFN-type fuzzy value. The matrix game is first converted to a fuzzy linear programming problem, which is converted to three different optimization problems, which are then solved to get the optimum value of the game. The proposed method has an edge over other method as this focuses only on matrix games with payoff element as symmetric trapezoidal fuzzy number, which might not always be the case. A numerical example is given to illustrate the method.

In this paper, we consider the scheduling problem in which the processing time of a job is a linear increasing function of its starting time and machine with availability constraints. The objective is to minimize the makespan. We first present a fully polynomial-time approximation scheme (FPTAS) for the case with a single machine. We then show that there exists no polynomial time approximation algorithm with a constant worst-case bound for the case with two identical machines unless $\mathrm{\text{P}}=\mathrm{\text{NP}}$.

This paper focuses on the berth allocation problem, which is to determine where and when the vessels to be loaded and unloaded at a terminal within a given planning horizon, with consideration of uncertain factors, mainly including the arrival and operation time of the calling vessels. Based on the concept of service level which is commonly used in the inventory system, a decision model is constructed to minimize the cost of baseline schedule, which includes delay cost and nonoptimal berthing location cost. According to the specific characteristics of the model, the upper and lower bounds are found. And due to the NP-hardness of the constructed model, an adaptive differential evolution is employed to solve the problem. Finally, extensive numerical experiments are conducted to test the performance of the proposed models and solution approaches.

In conventional data envelopment analysis (DEA), data are usually assumed to be non-negative with no specific bounds. However, many practical applications require some data, and thus their projections, to fall within certain limits. For example, percentage data such as the satisfactory rate cannot exceed 100% to make sense. This data characteristic is very likely to be violated under the assumption of constant returns to scale (CRS), due to its ray expansion property. In order to tackle this issue under CRS, a series of radial models are developed to constrain DEA projections within imposed bounds from the output side. Then efficient decision making units (DMUs) can be further discriminated simply by eliminating it from the reference set, avoiding the infeasibility problem existing in the VRS super-efficiency measures. The methodology is demonstrated with data consisting of 119 general acute care hospitals located in Pennsylvania, USA.

It is a common practice for online service providers to offer free experience service to attract new clients. However, providing experience service requires resources, which may negatively impact current service level and lead to customer turnover. Therefore, providers need to trade off between consequent benefits and costs. We consider a free experience service system where each arriving customer can start his service immediately without waiting, which is a typical situation of many online service platforms. A queueing model with infinite number of servers is employed to study such a service system. The closed forms of the expected numbers of informed and uninformed customers in steady-state are derived by solving nonhomogeneous linear partial differential equations. After that, the expected profit of the service provider is generated and maximized by determining the optimal price and service rates. Our numerical results suggest that with the increase of the market share and the serving cost for the informed customers, the service provider should lay more emphasis on offering the regular service for the informed customers.

By simultaneously considering both the players and the participation levels, we introduce two alternative definitions of the reduced games. Further, several axiomatizations of the unit-level-core are also obtained by applying related properties of consistency and its converse.