A number of quality loss functions, most recently the Taguchi loss function, have been developed to quantify the loss due to the deviation of product performance from the desired target value. All these loss functions assume the same loss at the specified specification limits. In many real life industrial applications, however, the losses at the two different specifications limits are often not the same. Further, current loss functions assume a product should be reworked or scrapped if product performance falls outside the specification limits. It is a common practice in many industries to replace a defective item rather than spending resources to repair it, especially if considerable amount of time is required. To rectify these two potential problems, this paper proposes more realistic quality loss functions for proper applications to real-world industrial problems. This paper also carries out a comparison studies of all the loss functions it considers.

In this paper we demonstrate the need for risk-adjusted approaches to planning expansion of livestock production. In particular, we illustrate that under exposure to risk, a portfolio of producers is needed where more efficient producers co-exist and cooperate with less efficient ones given that the latter are associated with lower, uncorrelated or even negatively correlated contingencies. This raises important issues of cooperation and risk sharing among diverse producers.

For large-scale practical allocation problems when information on the contingencies may be disperse, not analytically tractable, or be available on aggregate levels, we propose a downscaling procedure based on behavioral principles utilizing spatial risk preference structure. It allows for estimation of production allocation at required resolutions accounting for location specific risks and suitability constraints. The approach provides a tool for harmonization of data from various spatial levels. We applied the method in a case study of livestock production allocation in China to 2030.

Credit scoring has become a critical and challenging management science issue as the credit industry has been facing stiffer competition in recent years. Many classification methods have been suggested to tackle this problem in the literature. In this paper, we investigate the performance of various credit scoring models and the corresponding credit risk cost for three real-life credit scoring data sets. Besides the well-known classification algorithms (e.g. linear discriminant analysis, logistic regression, neural networks and k-nearest neighbor), we also investigate the suitability and performance of some recently proposed, advanced data mining techniques such as support vector machines (SVMs), classification and regression tree (CART), and multivariate adaptive regression splines (MARS). The performance is assessed by using the classification accuracy and cost of credit scoring errors. The experiment results show that SVM, MARS, logistic regression and neural networks yield a very good performance. However, CART and MARS’s explanatory capability outperforms the other methods.

Multi-agent technology has been applied extensively to many areas, including Decision Support Systems (DSS). However, the applications of multi-agent technology in DSS are still very preliminary. Most of the current agent frameworks, such as middle-agent-based or agent-facilitator-based frameworks, are basically agent-to-agent model. These agent-based frameworks often neglect the living environment for agents and they suffer from: (i) inability to adapt to the environment, (ii) inability to self-upgrade, and (iii) inefficiency in information acquisition. Here, we introduce a recently proposed multi-agent framework, namely Agent-based Open Connectivity for Decision Support Systems (AOCD). In this new framework, the communication and cooperation between agents are through a key component, the Matrix, which provides a virtual platform for agents. We use a unified Matrices framework to solve the bottleneck problem in the AOCD framework. Our experimental results based on different agent network topologies indicate that the hybrid topology presents superior performance compared with the centralised and decentralised topologies.

Fuzzy logic has difficulty producing valid answers in decision-making. Absent are theorems to prove that it works to produce results already known that are being estimated with judgments by transforming such judgments numerically. The numerical representation of judgments in the AHP is already fuzzy. Making fuzzy judgments more fuzzy does not lead to a better more valid outcome and it often leads to a worse one. The compatibility index of the AHP is used to illustrate how the answers obtained by fuzzifying AHP judgments do not produce better results than direct derivation of the principal eigenvector. Other authors who did experiments with given data in decision making quoted in the conclusions section of the paper, have observed that fuzzy sets gives the poorest answers among all methods used to derive best decisions.

The rate of return on capital is a key parameter in pension reform policy making. While evaluating pension reform, the method Feldstein proposed to measure the rate of return on capital is widely adopted. Here we calculate the rate of return on capital in China by this method. The calculation demonstrates that the rate of return on all the industrial enterprises is around 6.5 percent from 1996 to 2000, and the average rate of return on state-owned industrial enterprises is lower than the above figure by 1.5 percent during the same period. Finally, we draw a conclusion that the rate of return ranging from 5 to 7 percent is appropriate for the pension reform in China.

Ant-based text clustering is a promising technique that has attracted great research attention. This paper attempts to improve the standard ant-based text-clustering algorithm in two dimensions. On one hand, the ontology-based semantic similarity measure is used in conjunction with the traditional vector-space-model-based measure to provide more accurate assessment of the similarity between documents. On the other, the ant behavior model is modified to pursue better algorithmic performance. Especially, the ant movement rule is adjusted so as to direct a laden ant toward a dense area of the same type of items as the ant’s carrying item, and to direct an unladen ant toward an area that contains an item dissimilar with the surrounding items within its Moore neighborhood. Using WordNet as the base ontology for assessing the semantic similarity between documents, the proposed algorithm is tested with a sample set of documents excerpted from the Reuters-21578 corpus and the experiment results partly indicate that the proposed algorithm perform better than the standard ant-based text-clustering algorithm and the k-means algorithm.

This paper deals with a multi-level warehouse layout problem under fuzzy environment, in which different types of items need to be placed in a multi-level warehouse and the monthly demand of each item type and horizontal distance traveled by clamp track are treated as fuzzy variables. In order to minimize the total transportation cost, chance-constrained programming model is designed for the problem based on the credibility measure and then tabu search algorithm based on the fuzzy simulation is designed to solve the model. Some mathematical properties of the model are also discussed when the fuzzy variables are interval fuzzy numbers or trapezoidal fuzzy numbers. Finally, a numerical example is presented to show the efficiency of the algorithm.