INTERNATIONAL JOURNAL OF INDUSTRIAL ENGINEERING AND PRODUCTION MANAGEMENT (IJIE) (INTERNATIONAL JOURNAL OF ENGINEERING SCIENCE) (PERSIAN)
Year:2012 | Volume:23 | Issue:1|Papers Count:12

The intensive competition in e-Commerce causes effective methods for customer attraction of special importance. In this regard, the recommender systems in commercial websites can precisely determine customers' interests and needs, and offer them most suitable products and services. In this paper, a new model for recommender systems is proposed which segments the market and customers more efficiently, and then provides customers with better offers in two phases; customers are segmented based on demographic features such as age, gender, occupation and education in the first phase. The number of clusters are determined by means of the self-organizing map (SOM), and then clusters are created using K-means. In the second phase, association rules determine a valid map for each cluster which yields the most suitable offers to customers. To examine the efficiency of the proposed model in practice, it is used in an Iranian commercial website, and the results are analyzed.

The performance of an X-bar chart is usually studied under the assumption that the process standard deviation is well estimated and does not change. This is, of course, not always the case in practice and X-bar charts are not robust against errors in estimating the process standard deviation or changing standard deviation. In this paper, the use of an exponentially weighted moving average (EWMA) t chart with variable sampling intervals to monitor the process mean is discussed. We have determined the optimal control limits for the VSI EWMA t chart so that the chart has the desired robustness property against errors in estimating the process standard deviation or changing standard deviation. Performance of the proposed chart is compared with similar charts using the Markov chain approach and simulation studies.

Customer care and warranty options have become one of critical components of marketing. However, the warranty is known as an indication of quality, but Warranty in its basic assumptions is a procedure to handle the product cost tradeoff on its life cycle between producer, retailer and customer. Time – cost trade off from customer point of view versus reliability-cost tradeoff, as a producer/retailer concern on single component has been the main subject of most researches in this field. We have elaborated this problem to sophisticated products with more than one component with serial design basis. The model suggested in this paper redefines the existing approach of cost management to a revenue-based model. However, the optimal warranty period for a general demand distribution is considered in our paper, we have shown the result on case studies with exponential reliability function.

Banks as financial institutions must estimate the credit risk of their debtors. This is the basis of pricing a loan, determining appropriate interest rates and determining the mortgage required to each borrower. Since the continuity of bank activities largely depends on the amount of credit losses in a particular period, banks should consider the credit quality of their loan portfolio as a collection of debts.In this paper, the calculation of credit risk of corporates applying for loans has been investigated. Using fuzzy analytic hierarchy process, effective criteria for credit risk have been analyzed.The neural network is used to extract an open box model that describes the relationship between effective criteria and the credit risk of the companies who apply for a loan. Neural network model has been run with historical data. Observations have been based on 174 corporate who had taken out a loan from a major Iranian bank named Mellat (All loans had been made during 2005 to 2008).The output of the model can predict credit risk of a corporate by at least 84% accuracy.

In this paper, optimization of periodic inspection interval for a two-component system with failure dependency is presented. Failure of the first component is soft, namely, it does not cause the system stop, but it increases the system operating costs. The second component’s failure is hard, i.e. as soon as it occurs, the system stops operating. Any failure of the second component increases the first component’s failure rate. Failure of the first component is only detected if inspection is performed. Thus, the first component is periodically inspected and if found failed, it is perfectly repaired and it is restored to as good as new. Failure of the second component is detected as soon as it occurs. Since this failure causes the system stop, it is immediately replaced. It is assumed that the time for replacement or repaired is negligible. We model the first component’s failure as a non-homogeneous Poisson process (NHPP) with increasing failure rate and the second component’s failure as a homogeneous Poisson process (HPP) with constant failure rate. The objective is to find the optimal inspection interval for the first component such that the expected total cost per unit time is minimized. A simplified numerical example along with sensitivity analysis on cost parameters is given.

One kind of mid-term production system, aggregate production planning, identifies the optimum production plan for each production period. The goal of aggregate production planning is to forecast future demand swings. On the other hand, maintenance system identifies the proper time for preventive maintenance and restrains from break downs and reduces maintenance costs. Due to the importance of these two systems, in recent years, there has generated different models independently. The current research has proposed an integrated aggregate production planning model considering the time and costs of maintenance. This model indicates the optimum production size among the optimum time of preventive maintenance. As a final point, in order to check the reliability of the proposed model, an example has been examined. Results show that a considerable amount of cost has been saved by applying the model.

This study examines a multi-objective fuzzy simplex-genetic algorithm which was developed to predict bank legal customers financial performance. Prediction performance of the model was examined based on its ability to accurately identify credit default. Using available data from KESHVARZI bank over 2001-2006, debt ratio, operational ratio, and return on equity are selected as descriptive variables, and on the other side dependent variable was considered as a dummy variable. To training and validating the model, data were divided in to model (in-sample) and test (out-of-sample) sets. After running the algorithm, besides the sensitivity and specificity ratios, the key variable was specified.

The project scheduling problem is known as a NP-hard problem in literature. In this research, a resource constrained project scheduling problem which is known as a NP-Hard problem is considered. This problem has attracted many researchers during recent years. The aim of this problem is to determine the optimal starting times of activities considering both precedence and available resources constraints such that the total project completion time is minimized. In this paper a combination of discount based pricing policy and project scheduling is proposed, whereas in classical models it is assumed that price of required resources is fixed. To solve the proposed model, a hybrid algorithm based on two algorithms, i.e. genetic algorithm and variable neighborhood search is proposed. In this method, genetic algorithm as a main framework and variable neighborhood search as a new operator are designed. Moreover, since the parameter values of evolutionary algorithms have great influences on algorithm efficiency, to set the parameters of proposed algorithm a new statistical approach based on stepwise regression technique is devised. Computational results show the good performance of proposed approach with regard to the other methods.

Traditional project evaluation based on discounted cash flow analysis which ignores the upside potentials to an investment from managerial flexibility and innovation is not a suitable approach for evaluating high risk projects such as projects in oil industry. Nowadays, real options valuation approach that borrows ideas from financial options attracts the expert's attentions. In spite of the fact that experts have paid attention to this new method, applying this approach has some limitations. For applying this method successfully, we need to estimate some input parameters. One of the most important ones is volatility. volatility is a key parameter, but it is difficult to estimate. From a traditional investment viewpoint, volatility reduces project value because it increases its discount rate via a higher risk premium. The estimation of project volatility is very complicated since there is not a historical series of project values and most of projects are done for the first time and they are irreversible. In this article a method based on present value of future cash flows that we named profitability index and Monte Carlo simulation is proposed to estimate the volatility of projects. This method is applied to estimate the volatility of south pars gas field development phase 15 &16 as a case study.

The clustering problem under the criterion of minimum sum of squares is a non-convex and non-linear program, which possesses many locally optimal values, resulting that its solution often being stuck at locally optimal values and therefore cannot converge to global optima solution. In this paper, we introduce several new variation operators for the proposed hybrid genetic algorithm for the clustering problem. The novel mutation operator, called Clustering Regional Mutation, exchanges neighboring centers and a simple one-point crossover. The proposed algorithm identifies proper clustering. The experimental results are given to illustrate the effectiveness of the new genetic algorithm.

Because of the suitability of fuzzy numbers in representing uncertain values, ranking the fuzzy numbers has widely applications in different sciences. Many models are presented in field of ranking the fuzzy numbers that each one rank based on special criteria and features. The purpose of this paper is presenting a new method for ranking generalized fuzzy numbers based on some parameters such as membership degree, mean, standard deviation and parametric form of membership function. This paper introduce existing ranking models, comparing them the proposed model is presented .Considering special sets of fuzzy numbers, the advantages of proposed model is expressed. The main purpose of this paper is to present a reliable ranking method for generalized fuzzy numbers.