The methods of sensitivity analysis for linear programming can be classified in two types: sensitivity analysis using an optimal solution, and sensitivity analysis using an approximate optimal solution. As the methods of sensitivity analysis using an optimal solution, there are three sensitivity analysis methods: sensitivity analysis using an optimal basis, positive sensitivity analysis, and optimal partition sensitivity analysis. Since they may provide different characteristic regions under degeneracy, it is not easy to understand and apply the results of the three methods. In this paper, we propose a generalized sensitivity analysis that can integrate the three existing methods of sensitivity analysis. When a right-hand side or a cost coefficient is perturbed, the generalized sensitivity analysis gives different characteristic regions according to the controlling index set that denotes the set of variables allowed to have positive values in optimal solutions to the perturbed problem. We show that the three existing sensitivity analysis methods are special cases of the generalized sensitivity analysis, and present some properties of the generalized sensitivity analysis.

In this paper, we introduce a point-to-multipoint minimum cost flow problem with convex and demand splitting. A source node transmits the traffic along the tree that includes members of the point-to-multipoint connection. The traffic is replicated by the nodes only at branch points of the tree. In order to minimize the sum of arc costs, we assume that the traffic demand can be splitted and transmitted to destination nodes along different trees. If arc cost is linear, the problem would be a Steiner tree problem in networks eve though demand splitting is permitted. The problem would be applied in transmitting large volume of traffic from a serve to clients in Internet environments. Optimality conditions of the problem are presented in terms of fair tree routing. The proposed algorithm is a finite terminating algorithm for $\varepsilon$-optimal solution. convergence of the algorithm is obtained under monotonic condition and strict convexity of the cost function. Computational experiences are included.

In this study we consider a spares provisioning problem for repairable items in which a parts inventory system is incorporated. If a machine fails, a replacement part must be obtained at the parts inventory system before the failed machine enters the repair center. The inventory policy adopted at the parts inventory system is the (S, Q) policy. Operating times of the machine before failure, ordering lead times and repair times are assumed to follow a two-stage Coxian distribution. For this system, we develop an approximation method to obtain the performance measures such as steady state probabilities of the number of machines at each station and the probability that a part will wait at the parts inventory system. For the analysis of the proposed system, we model the system as a closed queueing network and analyze it using a product-form approximation method. A recursive technique as well as an iterative procedure is used to analyze the sub-network. Numerical tests show that the approximation method provides fairly good estimation of the performance measures of interest.

Estimation Problems of parameters of the failure process and the repair effect in repairable systems are considered. We propose estimation procedures in repairable systems without preventive maintenances. The failure process is modeled by a proportional age reduction model (Brown, Mahoney, Sivazlian [5]) which is able to consider both aging and repair effects. Maximum likelihood method is used to estimate the repair effect and parameters of intensity function simultaneously. simulations are performed to evaluate the accuracy of estimators. A numerical example is also presented.

This paper presents a linear programming based algorithm for pattern classification. Pattern classification is being considered to be critical in the area of artificial intelligence and business applications. Previous methods employing linear programming have been aimed at two-group discrimination with one or more linear discriminant functions. Therefore, there are some limitations in applying available linear programming formulations directly to general multi-class classification problems. The algorithm proposed in this manuscript is based on quadratic or polynomial discriminant functions, which allow more flexibility in covering the class regions in the N-dimensional space. The proposed algorithm is compared with other competitive methods of pattern classification in experimental results and is shown to be competitive enough for a general purpose classifier.

Symbiotic evolutionary algorithms, also called cooperative coevolutionary algorithms, are stochastic search algorithms that imitate the biological coevolution process through symbiotic interactions. In the algorithms, the fitness evaluation of an individual required first selecting symbiotic partners of the individual. Several partner selection strategies are provided. The goal of this study is to analyze how much partnering strategies can influence the performance of the algorithms. With two types of test-bed problems: the NKC model and the binary string covering problem, extensive experiments are carried out to compare the performance of partnering strategies, using the analysis of variance. The experimental results indicate that there does not exist statistically significant difference in their performance.

A knowledge-based system with fuzzy production rules is a representation of static knowledge of an expert. On the other hand, a real system such as the stock market is dynamic in nature. Therefore we need a strategy to reflect the dynamic nature of real system when we make inferences with a knowledge-based system. This paper proposes a strategy of dynamic inferencing for a knowledge-based system with fuzzy production rules. The strategy suggested in this paper applies weights of attributes of conditions of a rule in the knowledge-base. A degree of match(DM) between actual input information and a condition of a rule is represented by a value [0,1]. Weights of relative importance of attributes in a rule are obtained by AHP(Analytic Hierarcy Process) method. Then these weights are applied as exponents for the DM, and the DMs in a rule are combined, with MIN operator, into a single DM for the rule. In this way, overall DM for a rule changes depending on the importance of attributes of the rule. As a result, the dynamic nature of a real system can be incorporated in an inference with fuzzy production rules.

This work proposes a neural network approach to solve vehicle routing problems which have diverse application areas such as vehicle routing and robot programming. In solving these problems, classical mathematical approaches have many difficulties. In particular, it is almost impossible to implement a real-time vehicle routing with multiple vehicles. Recently, many researchers proposed methods to overcome the limitation by adopting heuristic algorithms, genetic algorithms, neural network techniques and others. The most basic model for path planning is the Travelling Salesman Problem(TSP) for a minimum distance path. We extend this for a problem with dynamic upcoming of new positions with multiple vehicles. In this paper, we propose an algorithm based on SOM(Self-Organization Map) to obtain a sub-optimal solution for a real-time vehicle routing problem. We develope a model of a generalized multiple TSP and suggest and efficient solving procedure.

This paper considers remanufacturing planning problems under deterministic environments. As increasing the environmental pressures in manufacturing, various methods for reducing wasted or postponing the time to be waste are considered. This paper considers remanufacturing planning problems on a single facility, where the wastes(or used products) are remanufactured to satisfy the given demand on the remanufactured products. The objective is to find the optimal remanufacturing and purchasing planning of the wastes which minimize total cost subject to satisfaction all the given demand on the remanufactured products. Two problems that the amount of wastes is a given constant or a decision variable are considered, respectively. For the problems, the extreme point solutions are characterized, and dynamic programming algorithms are developed with numerical examples.