Journal of Operational Research in its Applications
Year:2016 | Volume:13 | Issue:2 (49)|Papers Count:9

Designing bus transit networks is one of the most important issues in urban management. There are a large number of parameters influencing this design which are used to reach a set of goals such as improvement of the accessibility of citizens, maximum coverage of urban areas, reduction of waiting time and operational costs and number of transfer between line stops for a customer to reach its goal. Designing a Bus Transit Network is an NP-hard problem. This problem does not have optimal solution in large scale. The general way to design Bus Transit Network is as follows: Reducing the possible search space at the outset and then building the network-based on urban management priorities. In this paper, we proposed a new method to promote the design of a Bus Transit Network. Our approach is a statistical learning method which is designed through the help of statistical learning methods and combining them. In this study, knowledge of human experts from existing Bus Transit Networks is extracted. Then this knowledge is applied to reduce the design search space of a BTN in a small area of transits which have the necessary features to participate in BTNs and can be used in designing a Bus Transit Network Design Problem (BTNDP) or developing the current BTN. In this paper, we applied the Naive Bayesian, two regression-based methods, and the hybrid version of them to build the model. Evaluation of the learned model is based on accuracy, false positive and true positive criteria. The values of these criteria show high confidence of our approach. In this paper, we applied the Tehran Bus Transit Network as our data set.

Trust region subproblem (TRS), which is the problem of minimizing a quadratic function over a ball, plays a key role in solving unconstrained nonlinear optimization problems. Though TRS is not necessarily convex, there are efficient algorithms to solve it, particularly in large scale. Recently, extensions of TRS with extra linear constraints have received attention of several researchers. It has been shown that in the case where the linear constraints do not intersect within the ball, the optimal solution of the extended problem can be computed via solving a conic optimization problem. However, solving large-scale or even medium scale conic optimization problems are not practicable. In this paper, the extended trust region subproblem with two linear constraints without any assumptions on the constraints is considered. The latest algorithms for solving TRS and computing its local non-global minimizer that solve the problem via a generalized eigenvalue problem are used to solve the extended trust region subproblem. Finally, the efficiency of the proposed algorithm is evaluated on several randomly generated instances.

In scheduling, from both theoretical and practical points of view, a set of machines in parallel is a setting that is important. From the theoretical viewpoint, it is a generalization of the single machine scheduling problem. From the practical point of view, the occurrence of resources in parallel is common in real-world. When machines are computers, a parallel program is necessary because the members of the program are performed in a parallel fashion, and this performance is executed according to some precedence relationship. This paper shows the problem of allocating a number of non-identical tasks in a multi-processor or multicomputer system. The model assumes that the system consists of a number of identical processors, and only one task may be executed on a processor at a time. Moreover, all schedules and tasks are non-preemptive.

Acceptance sampling is one of the main parts of the statistical quality control. It is primarily used for the inspection of incoming or outgoing lots. Acceptance sampling procedures can be used in an acceptance control program to reach better quality with lower expenses, improved control, and increased efficiency. The aim of this paper is studying acceptance sampling based on non-parametric predictive inference in a fuzzy environment. In some cases, it may not be possible to define acceptance sampling parameters as crisp values. Especially in production environments, it may not be easy to define the number of parameters of conforming items or the size of the samples as crisp values. In these cases, these parameters can be expressed by linguistic variables. The fuzzy set theory can be used successfully to cope with the vagueness in these linguistic expressions for acceptance sampling based on non-parametric predictive inference. In other words, the aim of this paper is to present a new method titled fuzzy nonparametric predictive inference for single acceptance sampling plan.

In this paper, we investigate the backup undesirable 2-center location models on tree graphs. The aim is to obtain the best locations on the vertex set of the underlying tree for establishing two undesirable servers so that the expected value of the closest distance from the existing customers to the functioning facilities is maximized under the assumption that any facility may fail with a given probability and in this case the other active server must serve all the customers. The exact combinatorial algorithms with O(n2 log n) and O(n log n) time complexities are developed for obtaining the optimal solutions of two certain models, where n is the number of the vertices in the given tree graph.

In this paper, a new approach is presented for solving nonlinear systems of equations in which a derivative-free nonmonotone strategy is employed. Besides, the new approach is equipped with a filter technique. Using this concept, we store a lot of obtained trial points that are probably omitted by linear search algorithms. The new algorithm utilizes the information of existing points in the filter in order to obtain next points which causes a fast convergence rate in the algorithm. A specific nonmonotone technique is also used in the structure of the new approach which allows the algorithm to enjoy the nonmonotonicity feature from the first repetition. Under some standard assumptions, the global convergence property is established. Numerical results on some test problems show the efficiency of the proposed algorithm compared with some other existing methods in the literature.

In this paper we first define interval graphs and then consider the problem of finding a core of weighted interval trees. A core of an interval graph is a path containing some intervals of graph so that the sum of distances from all intervals to this path is minimized. We show that intervals on core of a tree cannot be interval non-maximal. Then an algorithm with O(n) time complexity is presented to find the core of interval trees.

In data envelopment analysis in which the comparison between a unit in relation to the antecedent set is used to detect the proportional inefficiency and providing an optimal model, detecting outliers is critical for achieving accurate results. In these deterministic frontier models, statistical tests are now applied to detect such outliers. This paper deals with introducing two statistical tests for detecting outliers in data envelopment analysis. In the presented methods, each observation is deleted from the sample once and the resulting DEA model is solved, leading to a distribution of efficiency estimates before and after the elimination. Based on the achieved distribution, two statistical tests are then designed to identify the potential outlier (s).We illustrate the method through a real data set. The method could be used in a first step before using any frontier estimation.