The difficulty on kansei evaluation stems from deciding the kansei specifications needed to realize a preference structure, namelyone of the human capabilities. In order to expand the analytical ability, it is of importance to construct a new model which will make clear the structure preferable to the population. In this paper, we attempt to analyze the kansei evaluation based on the least squares estimator and apply the preferential structure model to the data. This method is a special type of experimental design dealing with not absolute comparison, but relative comparison. If P_i is the proportion of the ith comparison subject, the preference structure model for the kansei evaluation with q comparison subjects is characterized by the constraints ΣP_i=1,P_1≧0. In this paper, the linear model z=Σα_iP_i and second-order model z=Σα_iP_i+ΣΣα_<ij>P_iP_j are used. We show computational examples of the preference structure model applied to the kansei evaluation of coffee beans to verify the effectiveness of the solution procedure, and clarify the applicability of the proposed method.

View full abstract

In recent years, many efforts have been made towards fuzzy scheduling problems where the uncertain processing time is treated as the fuzzy number. In this paper, we deal with unrelated parallel machine scheduling problems with fuzzy processing time and propose a fuzzy scheduling algorithm to minimize makespan, which can give more than one schedule as the solution. Our emphasis is particularly on the evaluating method of the solution schedules. A method is given to evaluate the solution schedules in order to choose one schedule as the best to be executed practically.

View full abstract

A New approach to canonical correlation analysis is proposed in which the Fuzzy c-Means clustering algorithm is simultaneously applied. Minimization of the within-group sum-of-squared-error and maximization of the canonical correlation coefficients yield memberships of fuzzy clusters and principal components of each cluster are analyzed from a complex of statistical variables. We analyze the opinions of youngsters by a questionnaire giving consideration to their family organization, aging, annuity and nursing. The results by the proposed method clearly show that opinions vary in accordance with the number of older family members living in the home environment.

View full abstract

Severe fluctuation in stock price is a factor in insecure economic activities such as bankruptcies and a fall in land prices. It's effective to find out the law of severe fluctuations of the sake of preventing fluctuations in stock price. Stock prices fluctuate as the result of interaction among decision-making agents and others. Therefore, it's difficult to express the factor for the fluctuation in stock prices. Therefore, we modeled an artificial stock market on a computer to investigate the influence of agents' characteristics on fluctuations in stock prices. Agent characteristics are defined as moment-to-moment changes in agent policies and an agent's basic disposition. We determine the law of severe fluctuations and propose how to prevent the fluctuations.

View full abstract

This paper deals with an economic ordering point model with a linear backorder ratio function which depends upon a backorder period. In the paper, demand during the lead time is assumed to be a random variable having a continuous distribution function. An average annual inventory cost (objective function), consisting of order cost, inventory holding cost, back-order cost and lost sales cost, is derived and then an iterative solution method for minimizing the objective function is presented. Finally, a numerical examination of how the reorder point and order quantity are related with the slope of the linear backorder ratio function is conducted.

View full abstract

It is required for the make-to-order company to design the production system which can meet the due date from customers as well as decrease work-in-process at a low level while keeping the machine utilization at a high level. This paper deals with the problem of order releasing in a job shop environments where external variation such as rush order cannot be negligible. Order release mechanisms are the set of rules to determine when matured orders to which materials and tools are available are dispatched to the shop floor to be processed. First, this paper explores the conditions that percent of order completed after their due date can be controlled within 5 and 1% respectively, assuming that rush order does not occur. Then we analyze the impact of rush order on the manufacturing performance in a dynamic situation that rush orders cannot be negligible. The following two performance measures are adopted in this study : Average earliness of all orders completed as inventory criteria and percent of orders completed after their due date as due date criteria. A simulation was conducted under various conditions of flow pattern of order which is characterized by the flow index, variation of operation time. The experimental results can provide management which order release type should be selected for improving the manufacturing performance.

View full abstract

Stochastic inventory models require information on the lead-time demand. However, information about the form of the probability distribution of the lead-time demand is often limited in practice. Sometimes all that is available is an educated guess of the mean and variance. There is a tendency to use the normal distribution under these conditions. However, the normal distribution does not provide the best protection against the occurrence of other distributions with the same mean and varianve. On the other hand, there is distribution-free approach that consists of finding the most unfavourable lead-time demand distribution for each decision variable and then minimizing over the decision variable. In this paper, we consider a fixed reorder quantity model in which allowable inventory at the end of a term is added to the model with a mixture of backorders and lost sales presented by Montgomery et al., and solve this model using the distribution-free approach.

View full abstract

In this paper, we propose an asymptotic Bayes optimal prediction algorithm for linear regression model, which reduces complexity in terms of calculation. In the field of industrial engineering, linear regression analyses are mainly applied to statistical quality control and demand forecasting, owing to effectiveness of control, prediction, analysis of structure, and so on. Recently, statistical model selection has been studied as a method of estimation for linear regression models, and applied to various problems of prediction. The statistical model selection is to select a particular model out of all candidates which include the true probabilistic model. The conventional criteria for model selection are F-value, FPE and information criteria; for example, AIC, BIC, and MDL. The mainly purposes of statistical model selection are to detect the true probability model, predict for future observations, and compress the data. Since statistical model selection has many applications, it has been studied not only in the field of statistics but also in various fields of science such as information theory, automatical control theory, and so on. In the case of estimation of the linear regression model by statistical model selection, generally, a particular model is selected by information criterion based on a previous observation from all candidates. In the linear regression analysis, the model class is a set of the combination of explanatory variables. However, in the case of prediction, it is not necessary to select a particular model. IN this case, the purpose of prediction is to acquire the accuracy estimator of the future observation. Therefore, previous studies using statistical model selection for prediction may be insufficient. On the other hand, the prediction method based on Bayes decision theory has been reported in various fields. In this method, predictions using the mixture model, which is mixing all candidates, are acquired as Bayes optimal solution, which minimizes the Bayesian mean square error. For this reason, we apply the mixture model for the linear regression models for prediction. We, at first, show that prediction by the mixture model is Bayes optimal prediction. However, it is difficult to strictly calculate the mixture probability because of the integration complexity on the parameter space. Therefore, we propose a new prediction method which removes the integration on account of reducing the complexity. Strictly speaking, we propose an asymptotic Bayes optimal prediction, which calculates the asymptotic posterior predictive distribution; i.e., mixture model. At last, we verify the effectiveness of the proposal through the simulation experiments.

View full abstract