Tomohiro Hayashida

A hybrid algorithm based on tabu search and ant colony optimization for k-minimum spanning tree problems

This paper considers an efficient approximate algorithm for solving k-minimum spanning tree problems which is one of the combinatorial optimization in networks. A new hybrid algorithm based on tabu search and ant colony optimization is provided. Results of numerical experiments show that the proposed method updates some of the best known values with very short time and that the proposed method provides a better performance with solution accuracy over existing algorithms.

Multiattribute utility analysis for policy selection and financing for the preservation of the forest

In this paper, we examine effective policies for financing and activities for the preservation of the forest on Mount Ryuoh in the city of Higashi-Hiroshima by multiattribute utility analysis. In multiattribute utility analysis, we deal with decision making problems with multiple attributes and select the most effective solution among several alternatives by deriving preference of the decision maker. Although in our decision making problem, the decision maker is a representative of a hypothetical nonprofit organization established for the preservation of the forest, the decision maker gives serious consideration to intentions of several groups of people receiving the benefit from the mountain, and then from this viewpoint, our problem can be interpreted as a group decision making problem.

An approximate solution method based on tabu search for k-minimum spanning tree problems

This paper considers a new tabu search-based approximate solution algorithm for k-minimum spanning tree problems. One of the features of the proposed algorithm is that it efficiently obtains local optimal solutions without applying minimum spanning tree algorithms. Numerical experimental results show that the proposed method provides a good performance especially for dense graphs in terms of solution accuracy over existing algorithms.

Sensitivity analysis incorporating fuzzy evaluation for scaling constants of multiattribute utility functions

A multiattribute utility function can be represented by a function of single-attribute utility functions if the decision maker’s preference satisfies additive independence or mutually utility independence. Additive independence is a preference condition stronger than mutually utility independence, and the multiattribute utility function is in the additive form if the former condition is satisfied, otherwise it is in the multiplicative form. In this paper, we propose a method for sensitivity analysis of multiattribute utility functions in multiplicative form, taking into account the imprecision of the decision maker’s judgment in the procedures for determining scaling constants (attribute weights).

Purchase and Transportation Planning for Food Retailing in Japan

In this paper, we consider purchase and transportation planning for food retailing in Japan, and formulate a linear programming problem where the profit of a food retailer is maximized. The food retailer deals with vegetables and fruits which are purchased at the central wholesale markets in several cities, and transports them by truck from each of the central wholesaler markets to the food retailers storehouse. After solving the linear programming problem and examining the optimal solution, we conduct sensitivity analysis by changing some parameters from managerial viewpoints. Finally, we discuss outsourcing of the purchase operations to a distributer.

Network formation and social reputation: a theoretical model and simulation analysis

In this paper, we deal with a network formation model where the utility of a player depends not only on the benefit from a network but also on the social reputation of the player. In the model, we examine the stability of networks and show the diversity of structures of the stable networks. Moreover, to complement the theoretical consideration, we perform simulation experiments for analysing the network formation. To do so, we develop a simulation system with artificial adaptive agents with a mechanism for decision making and learning based on neural networks and genetic algorithms.

A Hybrid Algorithm Based on Tabu Search and Ant Colony Optimization for k-Minimum Spanning Tree Problems

This paper considers an efficient approximate algorithm for solving k -minimum spanning tree problems which is one of the combinatorial optimization in networks. A new hybrid algorithm based on tabu search and ant colony optimization is provided. Results of numerical experiments show that the proposed method updates some of the best known values and that the proposed method provides a relatively better performance with solution accuracy over existing algorithms.

Multiattribute decision analysis using strict preference relations

To identify multiattribute utility functions under the assumption of the mutual utility independence, a decision maker must specify indifference points and subjective probabilities precisely. By relaxing this rigorous evaluation, multiattribute value models with incomplete information have been developed. In this paper, we present a new method finding the best alternative through strict preference relations derived from the decision maker by asking questions that are relatively easy to answer compared to the indifference questions. In our method, alternatives consistent with the derived strict preference relations are obtained by solving mathematical programming problems with constraints representing the preference relations, and eventually we can find the most preferred alternative by deriving additional preference relations.

Behavioural analysis in network formation using agent-based simulation systems

In the previous study on models of network formation and laboratory experiments, Berninghaus, Ehrhart, Ott and Vogt in 2007 showed that there are some differences between the real human behaviour in the laboratory experiments and the theoretical prediction based on the concept of Nash equilibrium; although a peripheral sponsored star network is theoretically proved to be strict Nash equilibrium, the experimental results showed that there were some players who deviated from the equilibrium strategies. As an efficient tool for explaining such differences, this paper provides a simulation system using artificial adaptive agents for behavioural analysis of the human subjects. The decision making and the learning mechanism of the agents are constructed on the basis of neural networks and genetic algorithms. Through the simulation analysis using the constructed system, it is shown that the differences between the experimental results and the theoretical prediction are explainable in terms of the long-term view and the adaptivity of human behaviour.

Computational Methods for Decentralized Two‐Level 0–1 Programming Problems through Distributed Genetic Algorithms

We consider two‐level programming problems in which there are one decision maker (the leader) at the upper level and two or more decision makers (the followers) at the lower level and decision variables of the leader and the followers are 0–1 variables. We assume that there is coordination among the followers while between the leader and the group of all the followers, there is no motivation to cooperate each other, and fuzzy goals for objective functions of the leader and the followers are introduced so as to take fuzziness of their judgments into consideration. The leader maximizes the degree of satisfaction (the value of the membership function) and the followers choose in concert in order to maximize a minimum among their degrees of satisfaction. We propose a modified computational method that solves problems related to the computational method based on the genetic algorithm (the existing method) for obtaining the Stackelberg solution. Specifically, the distributed genetic algorithm is introduced with...